<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD Journal Publishing DTD v2.3 20070202//EN" "journalpublishing.dtd">
<article xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" article-type="review-article" dtd-version="2.3" xml:lang="EN">
<front>
<journal-meta>
<journal-id journal-id-type="publisher-id">Front. Appl. Math. Stat.</journal-id>
<journal-title>Frontiers in Applied Mathematics and Statistics</journal-title>
<abbrev-journal-title abbrev-type="pubmed">Front. Appl. Math. Stat.</abbrev-journal-title>
<issn pub-type="epub">2297-4687</issn>
<publisher>
<publisher-name>Frontiers Media S.A.</publisher-name>
</publisher>
</journal-meta>
<article-meta>
<article-id pub-id-type="doi">10.3389/fams.2024.1380996</article-id>
<article-categories>
<subj-group subj-group-type="heading">
<subject>Applied Mathematics and Statistics</subject>
<subj-group>
<subject>Conceptual Analysis</subject>
</subj-group>
</subj-group>
</article-categories>
<title-group>
<article-title>Computational modeling of aging-related gene networks: a review</article-title>
</title-group>
<contrib-group>
<contrib contrib-type="author" corresp="yes"><name><surname>Freitas</surname> <given-names>Jos&#x00E9; Am&#x00E9;rico Nabuco Leva Ferreira</given-names></name><xref ref-type="corresp" rid="c001"><sup>&#x002A;</sup></xref>
<uri xlink:href="https://loop.frontiersin.org/people/2645958/overview"/>
<role content-type="https://credit.niso.org/contributor-roles/conceptualization/"/>
<role content-type="https://credit.niso.org/contributor-roles/funding-acquisition/"/>
<role content-type="https://credit.niso.org/contributor-roles/resources/"/>
<role content-type="https://credit.niso.org/contributor-roles/visualization/"/>
<role content-type="https://credit.niso.org/contributor-roles/writing-original-draft/"/>
<role content-type="https://credit.niso.org/contributor-roles/writing-review-editing/"/>
</contrib>
<contrib contrib-type="author" corresp="yes"><name><surname>Bischof</surname> <given-names>Oliver</given-names></name><xref ref-type="corresp" rid="c001"><sup>&#x002A;</sup></xref>
<uri xlink:href="https://loop.frontiersin.org/people/175501/overview"/>
<role content-type="https://credit.niso.org/contributor-roles/conceptualization/"/>
<role content-type="https://credit.niso.org/contributor-roles/funding-acquisition/"/>
<role content-type="https://credit.niso.org/contributor-roles/resources/"/>
<role content-type="https://credit.niso.org/contributor-roles/supervision/"/>
<role content-type="https://credit.niso.org/contributor-roles/writing-original-draft/"/>
<role content-type="https://credit.niso.org/contributor-roles/writing-review-editing/"/>
</contrib>
</contrib-group>
<aff><institution>IMRB, Mondor Institute for Biomedical Research, INSERM U955 &#x2013; Universit&#x00E9; Paris Est Cr&#x00E9;teil, UPEC, Facult&#x00E9; de M&#x00E9;decine de Cr&#x00E9;teil</institution>, <addr-line>Cr&#x00E9;teil</addr-line>, <country>France</country></aff>
<author-notes>
<fn fn-type="edited-by" id="fn0001">
<p>Edited by: Felix Sadyrbaev, University of Latvia, Latvia</p>
</fn>
<fn fn-type="edited-by" id="fn0002">
<p>Reviewed by: Jiajia Liu, University of Texas Health Science Center at Houston, United States</p>
<p>Amartya Singh, The State University of New Jersey, United States</p>
</fn>
<corresp id="c001">&#x002A;Correspondence: Jos&#x00E9; Am&#x00E9;rico Nabuco Leva Ferreira Freitas, <email>jose-americo.nabuco-leva@inserm.fr</email>; Oliver Bischof, <email>oliver.bischof@cnrs.fr</email></corresp>
</author-notes>
<pub-date pub-type="epub">
<day>28</day>
<month>05</month>
<year>2024</year>
</pub-date>
<pub-date pub-type="collection">
<year>2024</year>
</pub-date>
<volume>10</volume>
<elocation-id>1380996</elocation-id>
<history>
<date date-type="received">
<day>02</day>
<month>02</month>
<year>2024</year>
</date>
<date date-type="accepted">
<day>30</day>
<month>04</month>
<year>2024</year>
</date>
</history>
<permissions>
<copyright-statement>Copyright &#x00A9; 2024 Freitas and Bischof.</copyright-statement>
<copyright-year>2024</copyright-year>
<copyright-holder>Freitas and Bischof</copyright-holder>
<license xlink:href="http://creativecommons.org/licenses/by/4.0/">
<p>This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.</p>
</license>
</permissions>
<abstract>
<p>The aging process is a complex and multifaceted phenomenon affecting all living organisms. It involves a gradual deterioration of tissue and cellular function, leading to a higher risk of developing various age-related diseases (ARDs), including cancer, neurodegenerative, and cardiovascular diseases. The gene regulatory networks (GRNs) and their respective niches are crucial in determining the aging rate. Unveiling these GRNs holds promise for developing novel therapies and diagnostic tools to enhance healthspan and longevity. This review examines GRN modeling approaches in aging, encompassing differential equations, Boolean/fuzzy logic decision trees, Bayesian networks, mutual information, and regression clustering. These approaches provide nuanced insights into the intricate gene-protein interactions in aging, unveiling potential therapeutic targets and ARD biomarkers. Nevertheless, outstanding challenges persist, demanding more comprehensive datasets and advanced algorithms to comprehend and predict GRN behavior accurately. Despite these hurdles, identifying GRNs associated with aging bears immense potential and is poised to transform our comprehension of human health and aging. This review aspires to stimulate further research in aging, fostering the innovation of computational approaches for promoting healthspan and longevity.</p>
</abstract>
<kwd-group>
<kwd>aging</kwd>
<kwd>cellular senescence</kwd>
<kwd>gene-regulatory network</kwd>
<kwd>systems biology</kwd>
<kwd>computational modeling</kwd>
</kwd-group>
<counts>
<fig-count count="6"/>
<table-count count="2"/>
<equation-count count="0"/>
<ref-count count="139"/>
<page-count count="16"/>
<word-count count="13394"/>
</counts>
<custom-meta-wrap>
<custom-meta>
<meta-name>section-at-acceptance</meta-name>
<meta-value>Mathematical Biology</meta-value>
</custom-meta>
</custom-meta-wrap>
</article-meta>
</front>
<body>
<sec sec-type="intro" id="sec1">
<label>1</label>
<title>Introduction</title>
<p>Aging is a natural process that leads to a gradual decline in an individual&#x2019;s physiological function and reserve, diminishing their ability to survive and reproduce as they age. Its universal occurrence is a significant risk factor for developing several chronic diseases like cancer and cardiovascular or neurodegenerative diseases, ultimately leading to death bearing considerable individual and societal consequences and representing a growing global socio-economic burden. Consequently, substantial resources are dedicated to enhancing health span, focusing on the years an individual lives without chronic and debilitating diseases, in contrast to lifespan, which measures the total years from birth to death (<xref ref-type="bibr" rid="ref1">1</xref>, <xref ref-type="bibr" rid="ref2">2</xref>).</p>
<p>The quest to understand aging and identify molecular and cellular mechanisms promoting health span traces back to the pioneering work of Benjamin Gompertz (<xref ref-type="bibr" rid="ref3">3</xref>). Throughout history, reductionist science played a pivotal role in unraveling aging mechanisms, life span control, and age-related diseases. Notably, using <italic>Caenorhabditis elegans</italic> as a model organism, single-gene perturbation approaches yielded pivotal insights in aging research. The insulin-like signaling (ILS) pathway and germline signals, two major pathways, regulate lifespan across worms, flies, and mammals, with the transcription factor DAF-16/FOXO linking both. However, hundreds of other genes also contribute to longevity, complexifying the aging process (<xref ref-type="bibr" rid="ref4 ref5 ref6 ref7 ref8 ref9 ref10 ref11">4&#x2013;11</xref>).</p>
<p>L&#x00F3;pez-Ot&#x00ED;n et al. (<xref ref-type="bibr" rid="ref12">12</xref>, <xref ref-type="bibr" rid="ref13">13</xref>) condensed the intricate aging process and its underlying mechanisms into twelve hallmarks, offering a comprehensive framework to capture the essence of aging and comprehend its complexity. These interconnected hallmarks, including genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, disabled macroautophagy, deregulated nutrient-sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion, altered intercellular communication, chronic inflammation, and dysbiosis, contribute to the aging process, each representing a complex and multifaceted process influenced by various factors. Understanding these hallmarks is pivotal for developing interventions that delay aging and mitigate age-related diseases.</p>
<p>The complex nature of aging necessitates using <italic>in silico</italic> systems approaches that can offer a comprehensive framework to understand the fundamental principles governing gene regulatory interactions in aging organisms and predict aging genes. Various methods, such as Boolean networks, regression methods, correlation networks, functional networks, mutual information, ordinary and partial differential equation models, and neural networks, have been developed for depicting gene regulatory networks (GRNs), each tailored to specific input data and scientific questions.</p>
<p>Biological clocks, primarily utilizing epigenetic markers, accurately estimate biological age, providing more precise mortality risk estimation than chronological age. Contemporary techniques increasingly consider dynamic, nonlinear molecular interactions and leverage comprehensive published datasets. GRNs can benefit significantly from approaches involving higher orders of gene regulation, considering temporal changes in each molecular regulatory layer when integrating multi-OMICS datasets. Examples of these approaches, including studies by Schwanh&#x00E4;usser et al. (<xref ref-type="bibr" rid="ref14">14</xref>), La Manno et al. (<xref ref-type="bibr" rid="ref15">15</xref>), Hoffmann et al. (<xref ref-type="bibr" rid="ref16">16</xref>), and Badia-i-Mompel et al. (<xref ref-type="bibr" rid="ref17">17</xref>) have made substantial contributions to the study of gene regulatory networks, providing valuable insights into the molecular mechanisms underlying gene regulation.</p>
</sec>
<sec id="sec2">
<label>2</label>
<title>Mathematical models describing aging</title>
<p>The modeling of aging and age-related processes spans a wide array of methodologies applicable across various scales, from organismal to molecular levels (<xref ref-type="bibr" rid="ref2">2</xref>, <xref ref-type="bibr" rid="ref8">8</xref>, <xref ref-type="bibr" rid="ref18 ref19 ref20">18&#x2013;20</xref>). At the organismal scale, methodologies hinge on the foundational principles of molecular biology to deduce macroscopic properties. Conversely, molecular scale approaches operate on a data-driven basis, utilizing paradigms like Boolean logic, regression methods, information theory, neural networks, ordinary differential equations (ODEs), and partial differential equations (PDEs). It is crucial to emphasize that selecting an appropriate approach relies on the specific requirements of the study and the availability of resources. Therefore, a comprehensive understanding of the underlying principles becomes imperative when choosing and implementing these methodologies.</p>
<sec id="sec3">
<label>2.1</label>
<title>Macroscopic scale models</title>
<p>Santiago et al. (<xref ref-type="bibr" rid="ref2">2</xref>) offer an extensive review elucidating how asymmetry in cellular division shapes replicative aging in yeast, fostering a population with heightened robustness to fitness challenges. Complementing this, Kogan et al. (<xref ref-type="bibr" rid="ref21">21</xref>) introduce a mathematical model grounded in the topological properties of the genomic network within an individual, considering factors like network size (number of genes) and connectivity (average number of target genes per protein). Aligned with the Gompertz law, which posits an exponential increase in mortality with age across multiple species, this model implies that interventions targeting network stability prove more efficacious in extending lifespan than those aimed at reducing molecular stresses. This insight might explain the presence of long-lived organisms boasting extended health spans compared to closely related species.</p>
<p>From another perspective, Galvis et al. (<xref ref-type="bibr" rid="ref22">22</xref>) developed a model to explore the replicative exhaustion of fibroblasts intricately linked to cellular senescence. This model gauges cellular age/senescence based on the number of DNA replication cycles each cell undergoes. Throughout simulations, cells face growth arrest or apoptosis, becoming senescent after reaching 50 population doublings. Optimizing model parameters with experimental data, Galvis et al. capture the temporal activity of five senescence biomarkers (senescence-associated &#x03B2;-galactosidase, Ki-67 decrease, &#x03B3;H2AX, TUNEL, and population doublings). This underscores that population dynamics can be approximated using a relatively simple set of rules, shedding light on the mechanisms underpinning the cellular aging process.</p>
</sec>
<sec id="sec4">
<label>2.2</label>
<title>Boolean models</title>
<p>Boolean networks serve as a structural framework for describing GRNs, where each gene is a node with active or inactive states, and edges represent causal relationships. A gene&#x2019;s state hinges on its regulators, combined through Boolean functions incorporating OR, AND, or NOT operators. <xref ref-type="fig" rid="fig1">Figures 1A</xref>,<xref ref-type="fig" rid="fig1">B</xref> depict a Boolean network example and its associated equation, representing an architecture known as <italic>latch</italic>. The target gene T is regulated by a regulator R through an interaction modulated by gene M activity. A transcriptomic trajectory resulting from this GRN configuration is shown in <xref ref-type="fig" rid="fig1">Figure 1C</xref>.</p>
<fig position="float" id="fig1">
<label>Figure 1</label>
<caption>
<p>Boolean and fuzzy logic gene regulatory networks. <bold>(A)</bold> A Boolean gene regulatory network depicting a target gene T, activated by a regulator R. This interaction is controlled by the inhibitory modulator M. Standard symbols used to represent Boolean operators are shown. <bold>(B)</bold> Boolean logic equation representing the diagram depicted in <bold>(A)</bold>. T levels in the equation&#x2019;s right-hand side characterize a feedback loop, which, in this example, implements a memory effect in the network. <bold>(C)</bold> Time course T expression as a function of the inputs M and R. The inhibitory Modulator is initially inactivated, and R activation leads to T expression. Due to the feedback loop, T keeps high expression levels even after R is inactivated. Target T expression is reset to low values with the expression of the inhibitory modulator M. If M is expressed, activating R does not influence T levels. As soon as M is deactivated, T responds to R levels. <bold>(D)</bold> Equation describing T regulation using Fuzzy logic operators equivalent to those represented in panels <bold>(A,B)</bold>. <bold>(E)</bold> Membership function mapping gene expression to three distinct, overlapping classes. <bold>(F)</bold> Time course T expression as a function of M and R levels under a fuzzy logic framework. The fuzzy logic operators are applied to each track independently, and T class membership values are normalized at each instant to ensure they sum to 1. M is initially set to 80% Low and 20% Medium in this example. As the regulator changes from 100% Low to 10% Medium +90% High, T unnormalized levels rise to 10% Medium +90% High. T still presents a 20% Low membership level due to M&#x2019;s Low membership levels. As R levels transition at Time 20 from 10% Medium +90% High to 10% Low +90% Medium, T unnormalized levels for the Medium class rise, and unnormalized High levels are kept analogously to the Boolean logic case. When M is activated to 30% Medium +70% High, T unnormalized membership levels decrease to 70% Medium +30% High, with the membership class tending to Low. Slightly increasing R High membership at Time 40 does not affect T levels. Inactivating M by increasing its Low membership to 100% leads to a decrease in T Low membership.</p>
</caption>
<graphic xlink:href="fams-10-1380996-g001.tif"/>
</fig>
<p>Choo et al. (<xref ref-type="bibr" rid="ref23">23</xref>) leveraged a Boolean network model to pinpoint the minimal set of genes in the MAPK pathway that can cease cancer cell proliferation. Schwab et al. (<xref ref-type="bibr" rid="ref24">24</xref>, <xref ref-type="bibr" rid="ref25">25</xref>) evaluated the stability of aging signaling pathways using a Boolean network approach, assessing perturbations through mean normalized Hamming distances and reconstructing Boolean network ensembles from single-cell data, offering insights into the dynamics of aging in human hematopoietic stem cells. With the assumption that cells from the same patient are in equilibrium, i.e., that one cell can assume the transcriptional state exhibited by another in the same sample, they built pseudo-time trajectories for each patient. These trajectories were used to infer the regulators for each gene using Boolean functions. Since more than one function can lead to the same pseudo-time trajectory, they built a set of 1,000 networks per patient with distinct combinations of possible regulatory functions for each gene. They assessed the network&#x2019;s stability by changing a gene&#x2019;s state for each network and quantifying how many genes display a distinct Boolean state in the next time point compared to the unperturbed dynamic. Overall, the networks generated from hematopoietic stem cell data from young patients displayed a smaller discrepancy between perturbed and unperturbed dynamics than those generated from aged individuals.</p>
<p>Fuzzy logic extends Boolean networks by introducing more than two categories, like &#x2018;low,&#x2019; &#x2018;medium,&#x2019; and &#x2018;high&#x2019; (<xref ref-type="bibr" rid="ref26">26</xref>, <xref ref-type="bibr" rid="ref27">27</xref>). Gene expression levels map to these states through a &#x2018;membership function&#x2019; and regulatory interactions, executed by fuzzy logic operators (OR, AND, or NOT), output states with intermediary values for each category. The equation represented in <xref ref-type="fig" rid="fig1">Figure 1D</xref> demonstrates a Fuzzy logic extension for the Boolean operators shown in <xref ref-type="fig" rid="fig1">Figure 1A</xref>. By defining the membership for functions as portrayed in <xref ref-type="fig" rid="fig1">Figure 1E</xref>, we still observe how gene M influences target T dynamics, given the regulator R levels in <xref ref-type="fig" rid="fig1">Figure 1F</xref>. The obtained values are converted to one definite state using the inverse of the membership function to evaluate predictions. Aldridge et al. (<xref ref-type="bibr" rid="ref28">28</xref>) manually constructed a fuzzy network to model colon cancer response to membrane receptors, yielding less deviation from validation data than a discrete model with equivalent links. Poblete et al. (<xref ref-type="bibr" rid="ref29">29</xref>) employed a neural network to optimize parameters for a model with fuzzy membership functions describing gene regulation. Fuzzy networks offer enhanced flexibility and realism in modeling GRNs compared to Boolean networks.</p>
<p>In summary, Boolean and fuzzy networks prove potent for modeling GRNs, identifying gene sets to impede aging and ARDs, and deciphering intricate regulatory interactions. However, their suitability is limited to reduced-sized networks, and addressing the exponential growth of states with increasing nodes is essential. Moreover, designing specific solutions for categorizing continuous molecule concentrations into discrete bins is crucial for each application.</p>
</sec>
<sec id="sec5">
<label>2.3</label>
<title>Regression-based models</title>
<p>Regression analysis seeks optimal parameters to replicate a response using one or more independent variables (<xref ref-type="bibr" rid="ref30">30</xref>). This method aims to find a set of coefficients that minimize the difference between predicted and actual dependent variable values.</p>
<p>While regression analysis does not reveal underlying gene regulation mechanisms, it predicts gene behavior based on other gene expressions or phenotypic quantities like age or weight. For instance, Hackett et al. (<xref ref-type="bibr" rid="ref31">31</xref>) employed nonlinear regression to construct a yeast GRN by analyzing transcriptome responses to over 200 transcription factor (TF) activations. Machine learning algorithms identified regulatory interactions, unveiling new pathways and understanding their impact on gene expression patterns. Similarly, Li et al. (<xref ref-type="bibr" rid="ref32">32</xref>) developed a hybrid model integrating data from young and old individuals&#x2019; blood cells, using regression methods to identify factors contributing to changes in gene expression during aging.</p>
<p>Regression methods offer broad possibilities for analyzing quantities derived from measured data. Zhang et al. (<xref ref-type="bibr" rid="ref33">33</xref>) used linear regression to pinpoint age-related differentially expressed genes in 23 mouse tissues, defining global aging genes (GAGs). They determined GAG expression increases with age for each identified cell type, shedding light on genes and pathways in aging-related diseases.</p>
<p>Jordan and Miska (<xref ref-type="bibr" rid="ref34">34</xref>) analyzed <italic>C. elegans</italic> growth curves in a distinctive approach, constructing a two-dimensional nonlinear model with inferred parameters for each curve. This separated the dependency of each parameter on experimental variables (genotype and diet), highlighting at least two independent modulators regulating observed traits. This approach identified genetic factors in observed traits and revealed how environmental factors regulate them.</p>
<sec id="sec6">
<label>2.3.1</label>
<title>Biological clocks</title>
<p>The use of regression models in aging research to predict biological age, also called aging clocks, is a rapidly evolving field of study (<xref ref-type="bibr" rid="ref35">35</xref>). Hannum et al. (<xref ref-type="bibr" rid="ref36">36</xref>) pioneered using DNA methylation levels in blood cells to accurately predict chronological age, highlighting the difference between expected and actual age as an indicator of molecular damage. Advancing this, Horvath (<xref ref-type="bibr" rid="ref37">37</xref>) introduced a multi-tissue aging clock, proving valuable in primate studies. More recently, Levine et al. (<xref ref-type="bibr" rid="ref38">38</xref>) proposed a technique to forecast mortality risk using DNA methylation levels in blood cells. Unlike traditional age-based training, this method incorporates phenotypic parameters like liver and kidney metabolite concentration and white blood cell count measures into a regression model. In a further development, Belsky et al. (<xref ref-type="bibr" rid="ref39">39</xref>) introduced DunedinPACE (for Pace of Aging Calculated from the Epigenome), a next-generation DNA-methylation biomarker estimating the pace of aging.</p>
<p>Together, these approaches provide a more comprehensive evaluation of biological age and mortality risk, offering the potential for more effective interventions and treatments in ARDs.</p>
</sec>
<sec id="sec7">
<label>2.3.2</label>
<title>Multi-OMICS regression models</title>
<p>Partial Least Squares Regression (PLS) extends the linear regression model to handle multiple response variables, making it apt for analyzing multi-OMICS datasets (<xref ref-type="bibr" rid="ref40">40</xref>, <xref ref-type="bibr" rid="ref41">41</xref>). This method iteratively identifies components with the highest variance in each dataset, conducting a linear regression to predict molecule levels from the dataset considered a response variable. For instance, Rohart et al. (<xref ref-type="bibr" rid="ref41">41</xref>) utilized PLS to integrate mRNA, microRNA, and proteomics datasets from breast cancer samples, revealing markers distinguishing between different cancer subtypes. In another application, Th&#x00E9;venot et al. (<xref ref-type="bibr" rid="ref42">42</xref>) explored changes in the urine metabolome associated with aging, identifying 52 age-correlated metabolites, including increased carbohydrates and decreased acylcarnitines and nucleosides.</p>
<p>The Multi-Omics Factor Analysis (MOFA) framework integrates multi-OMICS datasets by pinpointing patterns in distinct molecular assays conducted on the same samples (<xref ref-type="bibr" rid="ref43">43</xref>, <xref ref-type="bibr" rid="ref44">44</xref>). These patterns, represented as latent factors, are linearly mapped to each assay. The framework extends to sparse datasets from single-cell assays [MOFA+; (<xref ref-type="bibr" rid="ref45">45</xref>)].</p>
<p>While regression-based methods are generally effective, they necessitate a sample size larger than the number of terms in the regression for a unique solution. In genomics studies, overcoming this challenge often involves assuming a sparse solution (<xref ref-type="bibr" rid="ref37">37</xref>) or representing it in terms of latent variables, grouping genes with similar profiles (<xref ref-type="bibr" rid="ref41">41</xref>).</p>
</sec>
</sec>
<sec id="sec8">
<label>2.4</label>
<title>Correlation networks</title>
<p>Correlation networks (CNs) play a pivotal role in OMICS studies, visually presenting pairwise correlations among molecular species. These networks, constructed from single-OMICS or multi-OMICS datasets, can be combined with graph theory metrics to highlight influential nodes (<xref ref-type="bibr" rid="ref18">18</xref>, <xref ref-type="bibr" rid="ref30">30</xref>, <xref ref-type="bibr" rid="ref46">46</xref>). In single-OMICS datasets, CNs reveal molecules sharing similar profiles, aiding in the selection of candidates for signaling pathways enrichment (<xref ref-type="bibr" rid="ref30">30</xref>, <xref ref-type="bibr" rid="ref41">41</xref>, <xref ref-type="bibr" rid="ref42">42</xref>, <xref ref-type="bibr" rid="ref47 ref48 ref49 ref50">47&#x2013;50</xref>). Our studies, exploring cells undergoing cellular senescence (CS) through metabolic, transcriptomic, and DNA accessibility datasets, employed CNs to profile molecular relationships (<xref ref-type="bibr" rid="ref51 ref52 ref53">51&#x2013;53</xref>).</p>
<p>In a study of cardiac failure due to high blood pressure, Froese et al. (<xref ref-type="bibr" rid="ref48">48</xref>) utilized CNs to analyze gene expression modules in cardiomyocytes, cardiac endothelial cells, and cardiac fibroblasts, which were used to identify transcription factor binding sequences enriched in co-expression modules. This provided novel regulatory molecules for hypertension therapies. TF motif binding, crucial from a systemic perspective, determines the dimensionality of the epigenetic landscape representing the phenotype (<xref ref-type="bibr" rid="ref34">34</xref>). In single-cell datasets, gene clustering aids in identifying cells with similar expression and associated gene markers. When applied to datasets collected from individuals with distinct ages, this procedure highlights physiological age-associated changes that lead to an increase in disease incidence (<xref ref-type="bibr" rid="ref54 ref55 ref56">54&#x2013;56</xref>).</p>
<p>Multi-OMICS CNs follow a similar construction, depicting molecules with similar accumulation profiles, even across different biochemical properties. Yu et al. (<xref ref-type="bibr" rid="ref57">57</xref>) identified highly correlated gene-protein pairs in cerebral endothelial cells, revealing their significance in an Alzheimer&#x2019;s disease model. Klaus et al. (<xref ref-type="bibr" rid="ref58">58</xref>) explored liver responses to a high-fat diet at metabolic and transcriptomic layers, unveiling metabolite-gene relationships with implications for Type 2 Diabetes.</p>
<p>In our studies on lung fibroblasts undergoing CS, we computed Spearman correlations between metabolomic and transcriptomic datasets under various stress sources, revealing central nodes like Glycerol Kinase (GK) (<xref ref-type="bibr" rid="ref53">53</xref>). Malod-Dognin et al. (<xref ref-type="bibr" rid="ref49">49</xref>) integrated multiple cancer samples, highlighting genes impacting <italic>in vitro</italic> cancer growth and stratifying patient data for distinct survival predictions.</p>
<p>Non-negative matrix factorization (NMF) also identifies latent factors associated with differentially accumulated molecules, considering that this association is mediated by non-negative coefficients (<xref ref-type="bibr" rid="ref40">40</xref>, <xref ref-type="bibr" rid="ref44">44</xref>). This premise ensures that the latent factors can be mapped to the dataset&#x2019;s features (i.e., genes, metabolites, or chromatin regions enriched with a histone mark), leading to more interpretable conclusions. The identified factors present similar coefficients for highly correlated molecules, simplifying downstream clustering techniques. Huang et al. (<xref ref-type="bibr" rid="ref59">59</xref>) analyzed mice&#x2019;s eye gene expression. This study identified a set of glaucoma-associated genes that could also discriminate ill mice samples from a second independent one. Zheng et al. (<xref ref-type="bibr" rid="ref60">60</xref>) applied an NMF-based approach to a list of genetic mutations and were able to stratify cancer types associated with each mutation set.</p>
<p>While correlation networks are versatile in portraying molecular features associated with experimental conditions, their representations may not correspond directly to mechanistic interactions, often a product of common modulation.</p>
</sec>
<sec id="sec9">
<label>2.5</label>
<title>Molecular interaction networks</title>
<p>Cellular metabolism involves molecules with distinct biochemical properties interacting through electrostatic binding sites and translocating between compartments via passive chemical gradients or active transport. Networks representing these interactions connect elements with shared biological functions, irrespective of their biochemical nature&#x2014;encompassing genes, proteins, histone marks, and enzymes catalyzing metabolic conversions.</p>
<p>In the field of molecular interaction networks, Chuang et al. (<xref ref-type="bibr" rid="ref18">18</xref>) explored their utility in identifying cancer biomarkers and cell fate decision drivers, integrating protein&#x2013;protein interactions (PPIs) and transcription factor (TF) binding site identification. Our studies delved into the dynamic chromatin landscape of senescent cells, integrating TF chromatin binding from ATAC-seq with histone marks detected by ChIP-seq, predicting target gene expression (<xref ref-type="bibr" rid="ref51">51</xref>, <xref ref-type="bibr" rid="ref52">52</xref>). Huang et al. (<xref ref-type="bibr" rid="ref61">61</xref>) compiled a database summarizing TF binding from scATAC-seq, enriched by scRNA-seq-detected gene expression in the same cell. Ren and Kuan (<xref ref-type="bibr" rid="ref62">62</xref>) benchmarked a multi-tissue biological age clock based on transcriptome data against DNA-methylation-based approaches.</p>
<p>While molecular interaction networks offer flexibility in conceptualizing multi-OMICS regulation, they often lack quantitative details about molecular reactions and may contain false-positive interactions. Integrating experimental data and complementary modeling approaches, such as regression methods (<xref ref-type="bibr" rid="ref16">16</xref>), becomes crucial to mitigate this.</p>
</sec>
<sec id="sec10">
<label>2.6</label>
<title>Mutual information</title>
<p>Mutual information approaches are statistical tools used to measure the predictability of one variable based on another. They quantify the degree of association and reveal dependencies between molecules. In simpler terms, if two molecules share high mutual information, knowing the levels of one molecule allows us to determine a narrow range that contains the value for the other molecule. This concept is illustrated by the GRN instance shown in <xref ref-type="fig" rid="fig2">Figure 2A</xref>, where, as previously, a gene M modulates the target T activation by regulator R. T gene levels are defined as shown in <xref ref-type="fig" rid="fig2">Figure 2B</xref>, which, under distinct M and R values, give rise to the multiple dynamics shown in <xref ref-type="fig" rid="fig2">Figure 2C</xref>. As expected, a high M value leads to no interaction between R and T, leading to a low mutual information between these two genes. As M decreases, the coupling between R and T increases, resulting in higher mutual information values. This phenomenon is evident in <xref ref-type="fig" rid="fig2">Figure 2D</xref>, where the cases associated with higher mutual information correspond to more precise T-level estimates by knowing that R values are contained in a small interval. Pioneering studies by Sachs et al. (<xref ref-type="bibr" rid="ref63">63</xref>), Chuang et al. (<xref ref-type="bibr" rid="ref18">18</xref>), and Linde et al. (<xref ref-type="bibr" rid="ref30">30</xref>) have used mutual information to uncover complex relationships among molecular entities.</p>
<fig position="float" id="fig2">
<label>Figure 2</label>
<caption>
<p>Mutual information strategies to identify modulatory interactions. <bold>(A)</bold> Gene regulatory network (GRN) showing how regulator gene R stimulates the activity of target gene T, which is controlled by the modulator gene M in a non-linear way. <bold>(B)</bold> Equation determining T levels as a function of R and M. T linearly increases as R is expressed, and the fraction of M determines the coefficient of this linear relationship compared to its maximum levels. <bold>(C)</bold> T gene expression changes as a function of R levels for various M activity levels. The Mutual Information Coefficient (MIC) between R and T for each M activity level is shown. <bold>(D)</bold> Comparison of T expression values for different R gene expression levels, sorted by M activity levels. The gray-filled boxplots represent the expression distribution of T for a specific range of R values shown in panel <bold>(C)</bold>. A higher T expression distribution for a particular set of R values suggests a stronger interaction between the regulator and the target. The strength of the interaction, known as the Maximum Information Coefficient (MIC), is dependent on the modulator level and hence highlights the need for non-linear methods for Gene Regulatory Network (GRN) inference.</p>
</caption>
<graphic xlink:href="fams-10-1380996-g002.tif"/>
</fig>
<p>The ARACNE (Algorithm for the Reconstruction of Accurate Cellular Networks) algorithm, a widely used method in GRN reconstruction, relies on mutual information. Developed by Lachmann et al. (<xref ref-type="bibr" rid="ref64">64</xref>), ARACNE finds applications in diverse research domains, from cardiomyopathy studies (<xref ref-type="bibr" rid="ref65">65</xref>) to investigations into female fertility (<xref ref-type="bibr" rid="ref66">66</xref>). Its ability to discern linear and nonlinear relationships renders ARACNE a robust tool for deciphering complex molecular dependencies.</p>
<p>Expanding on ARACNE, the MINDy (Modeling Interactions using Nonlinear Differential Equations) algorithm, conceived by Wang et al. (<xref ref-type="bibr" rid="ref67">67</xref>), identifies higher-order interactions between molecules. Analyzing mutual information in samples with varying modulator levels, MINDy unveils modulators, proving invaluable for pinpointing nonlinear dependencies and ascertaining interaction directions in GRNs.</p>
<p>Okada et al. (<xref ref-type="bibr" rid="ref68">68</xref>) have employed mutual information to craft a DNA methylation aging clock, gauging the correlation between DNA methylation patterns and age and offering a means to estimate an individual&#x2019;s biological age.</p>
<p>While mutual information approaches excel in detecting molecular dependencies, the computational intensity escalates when considering numerous molecular entities due to the requisite discrete classification of variable values.</p>
</sec>
<sec id="sec11">
<label>2.7</label>
<title>Ordinary differential equations</title>
<p>Ordinary Differential Equations (ODEs) serve as mathematical models for scrutinizing dynamic systems, particularly applicable in biology to capture gene expression changes in response to diverse stimuli. The versatility of ODEs allows the representation of nonlinear relationships among genes through various mathematical functions. <xref ref-type="fig" rid="fig3">Figure 3A</xref> portrays an example of GRN where the target gene T displays a self-regulatory loop that causes it to follow the expression values assumed by its regulator R. The ODE in <xref ref-type="fig" rid="fig3">Figure 3B</xref> represents this behavior, where the variation in T transcription is a parabolic function of T absolute gene expression. <xref ref-type="fig" rid="fig3">Figure 3C</xref> depicts this parabolic profile for distinct R values. This representation summarizes multiple GRN dynamical features, notably the potential presence of attractor states. Attractor states are characterized by a null velocity, i.e., the transcriptomic state is in equilibrium and will only change when the system is subjected to external stimuli (<xref ref-type="bibr" rid="ref19">19</xref>). These states can be stable if the system converges back to the attractor state when perturbed or unstable when the system diverges from the attractor state. An unstable attractor is also known as a &#x201C;tipping point.&#x201D; <xref ref-type="fig" rid="fig3">Figure 3D</xref> illustrates how distinct initial states converge to the stable attractor at T&#x2009;=&#x2009;R.</p>
<fig position="float" id="fig3">
<label>Figure 3</label>
<caption>
<p>Differential equations modeling nonlinear molecular interactions. <bold>(A)</bold> Gene regulatory network (GRN) depicting a logistic growth where the regulator R determines the stable expression level of its target T. The self-regulatory T edge ensures that its expression will follow R levels. <bold>(B)</bold> Differential equation describing the change in T expression concerning its value and R transcription levels. <bold>(C)</bold> Graphic representation of T expression rate for distinct R values. When T is expressed at the same levels as R, its transcription rate is zero, and the GRN is in equilibrium. These states are termed <italic>attractors</italic>. <bold>(D)</bold> Temporal T expression for the same R values depicted in the previous panel. The dynamics are shown for distinct initial T expression values T0. T transcription levels tend to R, even if T initially displays higher expression than R.</p>
</caption>
<graphic xlink:href="fams-10-1380996-g003.tif"/>
</fig>
<p>In our investigations, we harnessed ODE systems to depict the transcriptomic landscape of cellular senescence, analyzing 802-time gene expression profiles from TF knock-down (KD) experiments (<xref ref-type="bibr" rid="ref47">47</xref>). Through this approach, we also uncovered genes with potential hidden sources, which can be further explored in future iterations by integrating prior information, such as promoter-binding TFs or local epigenetic changes.</p>
<p>Konrath et al. (<xref ref-type="bibr" rid="ref69">69</xref>) constructed an ODE system to elucidate the inflammatory response triggered by DNA damage via PARP1 and ATM. Conducting sensitivity analysis by altering coefficients governing gene regulation, they revealed the potential efficacy of PARP1 inhibition in mitigating inflammatory responses during irradiation treatment of cancer cells.</p>
<p>Li et al. (<xref ref-type="bibr" rid="ref70">70</xref>) curated genes influencing aging-associated processes, formulating an ODE system grounded in sigmoid functions. Analyzing the system&#x2019;s probability, they identified three distinct attractors, one marked by accelerated aging with mTOR activation and another representing a slow aging state with upregulated SIRT1, AMPK, and Ulk1.</p>
<p>In a comprehensive discussion, Shin and Cho (<xref ref-type="bibr" rid="ref19">19</xref>) delved into the mathematical representation of systems and their potential landscapes to reverse cells undergoing malignant transformation. They proposed a two-step therapy involving rewiring the cellular internal network and destabilizing cancer-associated attractors in the first step. The second step focused on chronic microenvironment control to prevent cells from re-entering the attractor, which can be enhanced to leverage adaptive GRN motifs&#x2019; potential to reduce the need for prolonged microenvironment control in therapies.</p>
<sec id="sec12">
<label>2.7.1</label>
<title>Partial differential equations</title>
<p>ODEs are solved by integrating molecule evolution over one variable, typically time. At the same time, PDEs extend this integration to multiple variables, often spatial coordinates, allowing for incorporating spatial phenomena into simulations at the cellular and tissue levels (<xref ref-type="bibr" rid="ref8">8</xref>, <xref ref-type="bibr" rid="ref71">71</xref>).</p>
<p>Dyson et al. (<xref ref-type="bibr" rid="ref72">72</xref>) formulated a PDE system to model tumor growth, considering proliferating and quiescent cell populations, mutations, and cell migration influenced by extracellular nutrient gradients. In 2022, Son et al. expanded an ODE model of NF-kB dynamics, introducing a term for the spatial diffusion of TNF. Through experiments and simulations, they unveiled a gene expression response enabling cells to coordinate their reactions based on the dose, duration, and distance of tumor necrosis factor (TNF) signals.</p>
<p>Recent spatial transcriptomics studies have provided opportunities to explore tissue-level gene expression coordination and integrate spatiotemporal factors into model training. Kuppe et al. (<xref ref-type="bibr" rid="ref73">73</xref>) profiled spatial gene expression in millimeter-sized cardiac tissue sections after infarction from multiple human patients. Interestingly, they could identify ANKRD1 and NPPB upregulation at the injured area, accompanied by an increase in TGF&#x03B2; signaling-associated genes, detected by a pathway analysis performed in each spatial unit. Given that maximum spatial resolution can still be higher than the size of a cell, they combined this technique with scRNA-seq and were able to deconvolute cell types inside each spatial unit. This increased resolution led to the observation that ischemic samples have a higher proportion of macrophages expressing SPP1 and that these macrophages are associated with an increased myofibroblast presence in their surrounding area, supporting the hypothesis macrophages induce fibroblast differentiation into myofibroblasts. Yu et al. (<xref ref-type="bibr" rid="ref57">57</xref>) review multiple studies profiling spatial gene expression in cancer, highlighting the role of distinct cell types in tumor microenvironment and treatment response. Notably, colorectal cancer is associated with the colocalization of fibroblasts and SPP1-positive macrophages, leading to increased TGF&#x03B2; pathway activity. This mechanism leads to tissue fibrosis, eventually reducing immune infiltration and negatively impacting patient prognosis.</p>
<p>Despite a lack of studies performing GRN inference based on spatial transcriptomic data in age-related diseases, PDEs have been intensely studied in embryonic development and organ morphology. The <italic>Drosophila melanogaster</italic> embryo presents a concentration gradient for the <italic>bcd</italic> protein, activating downstream targets at distinct intensities and leading to increasingly sophisticated molecular patterns that define fly morphology (<xref ref-type="bibr" rid="ref74">74</xref>). Interestingly, Mousavi and Lobo (<xref ref-type="bibr" rid="ref75">75</xref>) have developed a tool that can identify GRN-originating spatial profiles from any arbitrary shape, given two genes regulate the network with gradients in two perpendicular dimensions. GRNs considering spatial information are usually modeled using reaction&#x2013;diffusion networks, which are similar to ODEs and contain a term associated with the spatial variation of regulator levels (<xref ref-type="bibr" rid="ref75">75</xref>, <xref ref-type="bibr" rid="ref76">76</xref>). This concept was formalized by Turing&#x2019;s (<xref ref-type="bibr" rid="ref77">77</xref>) early work on how simple rules generate complex spatial patterns, which, as discussed by Reinitz (<xref ref-type="bibr" rid="ref78">78</xref>) and Maini et al. (<xref ref-type="bibr" rid="ref79">79</xref>), has also gained renewed interest.</p>
<p>ODEs and PDEs prove powerful in modeling biological processes, understanding complex systems, and evaluating interventions through sensitivity analysis. Hidden sources analysis aids in identifying critical genes requiring attention, while identifying attractors in ODE systems offers insights into underlying mechanisms. Nevertheless, accurately representing GRNs via ODE inference demands substantial data and domain knowledge, and fitting noisy experimental data may lead to unstable dynamical systems.</p>
</sec>
</sec>
<sec id="sec13">
<label>2.8</label>
<title>Artificial networks</title>
<p>Artificial neural networks, a subset of nonlinear machine learning techniques (<xref ref-type="bibr" rid="ref17">17</xref>, <xref ref-type="bibr" rid="ref27">27</xref>, <xref ref-type="bibr" rid="ref30">30</xref>, <xref ref-type="bibr" rid="ref80">80</xref>), excel in describing a wide range of processes with superior accuracy compared to linear models, although they require more data for precise identification. <xref ref-type="fig" rid="fig4">Figure 4A</xref> depicts a neural network implementing a high-order polynomial function with one input. It contains one hidden layer with 15 neurons. Each of these neurons outputs a value based on the function shown in <xref ref-type="fig" rid="fig4">Figure 4B</xref>, known as the &#x201C;activation function.&#x201D; In this example, the activation function is a Rectified Linear Unit (ReLU), which acts as an identity map for positive values and returns zero for all negative input values. Before being applied to the activation function, the input is linearly transformed by the weights w<sub>i</sub>, which are the parameters to be optimized during the network training. <xref ref-type="fig" rid="fig4">Figure 4C</xref> shows the impact of each neuron (blue curves) in the output and the cumulative effect of all previous neurons to compute the target polynomial function. The predicted and target functions are depicted in <xref ref-type="fig" rid="fig4">Figure 4D</xref>. Deep learning neural networks (DNN), consisting of multiple intermediate layers, have the property of modeling any mathematical function and revolutionizing image analysis and object recognition, showcasing unprecedented accuracy (<xref ref-type="bibr" rid="ref81">81</xref>). In genomics and aging studies, DNNs have proven valuable for inferring GRNs, deconvoluting cell types, and constructing biological clocks.</p>
<fig position="float" id="fig4">
<label>Figure 4</label>
<caption>
<p>Neural networks as universal approximators. <bold>(A)</bold> Description of a neural network trained to approximate a target function. The network has one input, one hidden layer with 15 neurons, and one output. The black edges in the network represent the weights that determine the activation of each neuron based on the input value. The blue edges represent the independent terms that lead to the basal activation of each neuron, also known as biases. Each neuron applies an operation, determined by an activation function, to its input. <bold>(B)</bold> The Rectified Linear Unit (ReLU) function is the activation function for this example. <bold>(C)</bold> Contribution of each neuron to the final model prediction. As more neurons are considered, the model prediction approaches the target function. <bold>(D)</bold> Neural network model output and target function comparison, with a Root Means Squared Error (RMSE) of 0.057.</p>
</caption>
<graphic xlink:href="fams-10-1380996-g004.tif"/>
</fig>
<p>Cancer GRN inference has been explored using DNNs, as seen in studies by Wysocka et al. (<xref ref-type="bibr" rid="ref20">20</xref>) and Langeman et al. (<xref ref-type="bibr" rid="ref80">80</xref>) on leukemia regulation. Autoencoder architecture, employed in Alzheimer&#x2019;s research (<xref ref-type="bibr" rid="ref82">82</xref>), aids in determining cell proportions. Jordan &#x0026; Miska (<xref ref-type="bibr" rid="ref34">34</xref>) applied autoencoders to nonlinearly separate <italic>C. elegans</italic> growth parameters based on diet or genotype. Qiu et al. (<xref ref-type="bibr" rid="ref83">83</xref>) used decision trees to estimate biological age with higher interpretability than DNNs. Holzscheck et al. (<xref ref-type="bibr" rid="ref84">84</xref>) designed a biological clock using a DNN with parallel subnetworks mimicking biological signaling, allowing the identification of active neurons and their functions during age inference. Tao et al. (<xref ref-type="bibr" rid="ref85">85</xref>) benchmarked multiple artificial intelligence methods to classify proliferation and senescent cells and identified six transcriptomic signatures associated with distinct CS identities.</p>
<p>Despite the effectiveness of artificial intelligence (AI) methods, challenges persist in situations requiring high safety levels, like healthcare or autonomous driving. To address this, paradigms like Conformal Prediction (<xref ref-type="bibr" rid="ref86">86</xref>) have emerged, providing metrics to quantify the certainty of model predictions. Conformal prediction offers insights into the reliability of AI models for specific use cases, emphasizing the importance of understanding prediction variations associated with confidence levels.</p>
</sec>
</sec>
<sec id="sec14">
<label>3</label>
<title>Perspectives</title>
<sec id="sec15">
<label>3.1</label>
<title>Potentially disruptive datasets</title>
<p>The field of systems biology and aging is poised to benefit significantly from advancements in AI, akin to the transformative impact witnessed in image processing breakthroughs. The ImageNet dataset revolutionized image processing by providing over 10 million images with a standardized set of labels (<xref ref-type="bibr" rid="ref81">81</xref>); the GPT-3 large language model was trained using an unprecedented number of parameters and has exhibited superior accuracy across various language tasks compared to previous versions without annotated data (<xref ref-type="bibr" rid="ref108">108</xref>). Therefore, achieving comparable breakthroughs in systems biology and aging necessitates the development of comprehensive expert-annotated datasets.</p>
<p>Fortuitously, global collaborative efforts have facilitated the integration of numerous high-throughput datasets into a unified framework (<xref ref-type="table" rid="tab1">Table 1</xref>). Uyar et al. (<xref ref-type="bibr" rid="ref55">55</xref>) enumerate studies characterizing age-related transcriptomic changes at the single-cell level in multiple organs. Furthermore, larger-scale initiatives are actively underway to delineate multi-tissue age-related pathophysiological changes. These initiatives involve databases consolidating raw data from projects investigating mechanisms related to the same diseases and international consortiums dedicated to comprehending and mapping the molecular features of both healthy individuals and pathological molecular trajectories. These initiatives significantly advance our understanding of the fundamental mechanisms underlying aging and disease by aggregating findings from diverse studies into a cohesive format.</p>
<table-wrap position="float" id="tab1">
<label>Table 1</label>
<caption>
<p>Large-scale initiatives mapping human healthy and pathological molecular states.</p>
</caption>
<table frame="hsides" rules="groups">
<thead>
<tr>
<th align="left" valign="top">Name</th>
<th align="left" valign="top">Reference</th>
<th align="left" valign="top">Description</th>
</tr>
</thead>
<tbody>
<tr>
<td align="left" valign="top">Pan-Cancer Atlas</td>
<td align="left" valign="top">Sanchez-Vega, et al. (<xref ref-type="bibr" rid="ref87">87</xref>); Ding et al. (<xref ref-type="bibr" rid="ref88">88</xref>); Hoadley et al. (<xref ref-type="bibr" rid="ref89">89</xref>)</td>
<td align="left" valign="top">Large-scale collaboration harmonizing the analysis of the Cancer Genome Atlas database divided into three categories: cell-of-origin patterns, oncogenic processes, and signaling pathways</td>
</tr>
<tr>
<td align="left" valign="top">Type 2 Diabetes Knowledge Portal (T2DKP)</td>
<td align="left" valign="top">Costanzo et al. (<xref ref-type="bibr" rid="ref90">90</xref>)</td>
<td align="left" valign="top">A knowledge base integrating transcriptome, phenotype, and genomic variant information for genes associated with T2D</td>
</tr>
<tr>
<td align="left" valign="top">Cardiovascular Disease Knowledge Portal (CVDKP)</td>
<td align="left" valign="top">Costanzo et al. (<xref ref-type="bibr" rid="ref91">91</xref>)</td>
<td align="left" valign="top">A knowledge base integrating transcriptome, phenotype, and genomic variant information for genes associated with cardiovascular diseases</td>
</tr>
<tr>
<td align="left" valign="top">Open Genes</td>
<td align="left" valign="top">Rafikova et al. (<xref ref-type="bibr" rid="ref92">92</xref>)</td>
<td align="left" valign="top">A set of 2,402 genes associated with aging</td>
</tr>
<tr>
<td align="left" valign="top">Human Ageing Genomic Resources (HAGR)</td>
<td align="left" valign="top">Tacutu et al. (<xref ref-type="bibr" rid="ref11">11</xref>)</td>
<td align="left" valign="top">A collection of aging-associated databases, including genes associated with aging, genetic association longevity studies, and longevity-associated chemical compounds</td>
</tr>
<tr>
<td align="left" valign="top">scRNA-tools</td>
<td align="left" valign="top">Zappia et al. (<xref ref-type="bibr" rid="ref93">93</xref>)</td>
<td align="left" valign="top">Database of tools used in scRNA-seq analysis</td>
</tr>
<tr>
<td align="left" valign="top">Lifetime Initiative</td>
<td align="left" valign="top">Rajewsky et al., (<xref ref-type="bibr" rid="ref94">94</xref>)</td>
<td align="left" valign="top">International collaboration dedicated to characterizing spatiotemporal cellular trajectories of healthy and diseased tissues and organs</td>
</tr>
<tr>
<td align="left" valign="top">Human Cell Atlas</td>
<td align="left" valign="top">Regev et al. (<xref ref-type="bibr" rid="ref95">95</xref>)</td>
<td align="left" valign="top">International collaboration dedicated to mapping every human cell type</td>
</tr>
<tr>
<td align="left" valign="top">Human Protein Atlas (HPA)</td>
<td align="left" valign="top">Thul and Lindskog (<xref ref-type="bibr" rid="ref96">96</xref>)</td>
<td align="left" valign="top">Tissue-specific protein expression patterns determined by immunohistochemistry and immunocytochemistry imaging</td>
</tr>
<tr>
<td align="left" valign="top">HPA Single Cell Type Atlas</td>
<td align="left" valign="top">Karlsson et al. (<xref ref-type="bibr" rid="ref97">97</xref>)</td>
<td align="left" valign="top">scRNA-seq and spatial antibody-based imaging integration</td>
</tr>
<tr>
<td align="left" valign="top">ENCODE</td>
<td align="left" valign="top">Roadmap Epigenomics Consortium et al. (<xref ref-type="bibr" rid="ref98">98</xref>)</td>
<td align="left" valign="top">Epigenetic data (histone marks, DNA methylation, and DNA accessibility) for multiple cell types</td>
</tr>
</tbody>
</table>
</table-wrap>
<p>Exploring different species holds promise for therapeutic development (<xref ref-type="table" rid="tab2">Table 2</xref>). Tejada-Martinez et al. (<xref ref-type="bibr" rid="ref109">109</xref>) uncovered 71 duplicated tumor-suppressor genes in cetaceans, potentially explaining their extended lifespan. These genes offer a valuable avenue for investigating cancer growth reduction and delaying tumor relapse in humans. Wang et al. (<xref ref-type="bibr" rid="ref56">56</xref>) probed scRNA-seq data across mice, zebrafish, and flies, revealing age-associated immune response activation and diminished mitochondrial activity. Mah and Dunn (<xref ref-type="bibr" rid="ref110">110</xref>) constructed a multi-species phylogenetic tree from scRNA-seq data, pinpointing common, evolutionarily conserved pathways and quantifying transcriptomic response similarity across species. Intriguingly, Jiang et al. (<xref ref-type="bibr" rid="ref102">102</xref>) noted a significant correlation between the transcriptomes of zebrafish swim bladder epithelial cells and mice alveolar type 1 cells, in line with their specialized functions in gas exchange&#x2014;<xref ref-type="table" rid="tab2">Table 2</xref> catalogs single-cell datasets from diverse species, offering potential targets for delaying age-related pathologies. Additional pertinent studies by Wang et al. (<xref ref-type="bibr" rid="ref111">111</xref>) and Fei et al. (<xref ref-type="bibr" rid="ref101">101</xref>) contribute to this comprehensive exploration.</p>
<table-wrap position="float" id="tab2">
<label>Table 2</label>
<caption>
<p>Single-cell OMICS datasets describing multi-organ age-associated molecular changes.</p>
</caption>
<table frame="hsides" rules="groups">
<thead>
<tr>
<th align="left" valign="top">Species</th>
<th align="left" valign="top">References</th>
<th align="left" valign="top">Description</th>
</tr>
</thead>
<tbody>
<tr>
<td align="left" valign="top">Human</td>
<td align="left" valign="top">The Tabula Sapiens Consortium et al. (<xref ref-type="bibr" rid="ref99">99</xref>)</td>
<td align="left" valign="top">Multi-organ scRNA-seq</td>
</tr>
<tr>
<td align="left" valign="top">Human, mouse</td>
<td align="left" valign="top">Huang et al. (<xref ref-type="bibr" rid="ref61">61</xref>)</td>
<td align="left" valign="top">TF-gene target pairs inferred from scATAC-seq and scRNA-seq integration</td>
</tr>
<tr>
<td align="left" valign="top">Mouse, fly, zebrafish</td>
<td align="left" valign="top">Wang et al. (<xref ref-type="bibr" rid="ref56">56</xref>)</td>
<td align="left" valign="top">Multi-species scRNA-seq integration over age</td>
</tr>
<tr>
<td align="left" valign="top">Mouse</td>
<td align="left" valign="top">The Tabula Muris Consortium (<xref ref-type="bibr" rid="ref54">54</xref>)</td>
<td align="left" valign="top">Multi-organ scRNA-seq at distinct ages</td>
</tr>
<tr>
<td align="left" valign="top">Mouse</td>
<td align="left" valign="top">Cusanovich et al. (<xref ref-type="bibr" rid="ref100">100</xref>)</td>
<td align="left" valign="top">Multi-organ scATAC-seq</td>
</tr>
<tr>
<td align="left" valign="top">Mouse</td>
<td align="left" valign="top">Fei et al. (<xref ref-type="bibr" rid="ref101">101</xref>)</td>
<td align="left" valign="top">Multi-organ scRNA-seq at different ages, including embryonic stages</td>
</tr>
<tr>
<td align="left" valign="top">Zebrafish</td>
<td align="left" valign="top">Jiang et al. (<xref ref-type="bibr" rid="ref102">102</xref>)</td>
<td align="left" valign="top">Multi-organ scRNA-seq at distinct ages, including embryonic stages; scRNA-seq characterization of fin regeneration</td>
</tr>
<tr>
<td align="left" valign="top">Zebrafish</td>
<td align="left" valign="top">Wagner et al. (<xref ref-type="bibr" rid="ref103">103</xref>)</td>
<td align="left" valign="top">Multi-organ scRNA-seq from distinct embryonic stages</td>
</tr>
<tr>
<td align="left" valign="top">Worm</td>
<td align="left" valign="top">Roux et al. (<xref ref-type="bibr" rid="ref104">104</xref>)</td>
<td align="left" valign="top">Multi-organ scRNA-seq at distinct ages</td>
</tr>
<tr>
<td align="left" valign="top">Worm</td>
<td align="left" valign="top">Packer et al. (<xref ref-type="bibr" rid="ref105">105</xref>)</td>
<td align="left" valign="top">Multi-organ scRNA-seq from different embryonic stages</td>
</tr>
<tr>
<td align="left" valign="top">Hydra</td>
<td align="left" valign="top">Siebert et al. (<xref ref-type="bibr" rid="ref106">106</xref>)</td>
<td align="left" valign="top">Multi-organ scRNA-seq from a stem cell-rich, potentially immortal organism</td>
</tr>
<tr>
<td align="left" valign="top">Planaria</td>
<td align="left" valign="top">Plass et al. (<xref ref-type="bibr" rid="ref107">107</xref>)</td>
<td align="left" valign="top">Multi-organ scRNA-seq from a stem cell-rich, potentially immortal organism</td>
</tr>
</tbody>
</table>
</table-wrap>
</sec>
<sec id="sec16">
<label>3.2</label>
<title>Potentially disruptive methods</title>
<p>Aging research is constantly evolving, and new data analysis methods are emerging. In addition to producing high-quality, comprehensive datasets, deep neural networks (DNNs) have achieved unparalleled accuracy. While the field is still under active research, DNN architectures can better estimate data from different domains with specific distributions (<xref ref-type="bibr" rid="ref112">112</xref>, <xref ref-type="bibr" rid="ref113">113</xref>). DNN architecture is critical as it can bridge the gap between formal models based on fixed principles and virtually unbounded approaches. The literature provides a detailed description of the properties associated with each architecture (<xref ref-type="bibr" rid="ref114">114</xref>). We have highlighted specific DNN architectures that have the potential to enhance our understanding of gene regulatory networks (GRN) that regulate age-associated processes. Combining these architectures with the datasets in <xref ref-type="table" rid="tab2">Table 2</xref> can significantly improve our understanding of aging.</p>
<sec id="sec17">
<label>3.2.1</label>
<title>Dimensionality reduction: autoencoders and Kohonen networks</title>
<p>Autoencoders are a class of deep neural networks that utilize two sub-networks connected in series: an encoder and a decoder. The encoder maps the input data to an intermediate layer with fewer neurons. At the same time, the decoder reconstructs the input data from the information available in this layer with reduced degrees of freedom. As a result, the network generates a low-dimensional data representation. Due to the non-linear activation function of neurons, this representation can also map non-linear functions. Previous research has shown that this lower-dimensional representation can represent transcriptional profiles of distinct cell types in a biological sample, which can be used in other samples to estimate subpopulation proportions (<xref ref-type="bibr" rid="ref82">82</xref>). This tool can also separate potential sources of transcriptional variability, which can be later mapped to genetic or environmental conditions (<xref ref-type="bibr" rid="ref34">34</xref>).</p>
<p>Another technique that identifies a low-dimensional data representation is the Kohonen Network, known as &#x201C;self-organizing maps&#x201D; (<xref ref-type="bibr" rid="ref115">115</xref>). Despite not following the conventional neural network representation, Kohonen Networks associate each input with a cell in a grid so that inputs with closer dynamics will be closer to the map. In addition to determining a low-dimensional data representation, it effectively functions as a clustering technique.</p>
</sec>
<sec id="sec18">
<label>3.2.2</label>
<title>Residual neural networks</title>
<p>Residual networks, also known as ResNets, are a type of neural network that includes &#x201C;shortcut connections&#x201D; that start at neurons in one layer and end at neurons in layers several steps ahead (<xref ref-type="bibr" rid="ref116">116</xref>). This unique feature allows intermediate layers to model only the difference between the output of the layer at the beginning of the shortcut and the expected training output. This leads to greater network robustness and faster training convergence, which allows networks to stack hundreds of layers. Recent studies have shown that ResNets are effective in various applications. For example, Cao et al. (<xref ref-type="bibr" rid="ref117">117</xref>) used a DNN with a ResNet submodule to detect and localize myocardial infarction sites from a 12-lead electrocardiogram signal. Additionally, Habijan et al. (<xref ref-type="bibr" rid="ref118">118</xref>) used ResNet as an autoencoder network to design intermediary layers for segmenting heart chambers from computer tomography (CT) and magnetic resonance imaging (MRI) data. This led to higher similarity between predictions and expert-annotated data.</p>
</sec>
<sec id="sec19">
<label>3.2.3</label>
<title>Transformers</title>
<p>Transformers have significantly impacted artificial intelligence due to their application in large language models. However, their implementation has raised ethical and cultural concerns (<xref ref-type="bibr" rid="ref119">119</xref>). Initially introduced by Vaswani et al. (<xref ref-type="bibr" rid="ref120">120</xref>) for language translation, transformers are now widely used in text generation to predict the next word in a sentence. The transformer architecture&#x2019;s main innovation is the &#x201C;attention mechanism,&#x201D; which enables it to associate words that refer to the same concept. Three matrices, Key (K), Query (Q), and Value (V), are used for this mechanism, with each word in the input associated with a K matrix. This matrix calculates the KQ product for each word in the sequence, with Q depending on the meaning of each input word. As a result, words referring to the same idea will have higher KQ product values. The V matrix determines the most probable word to continue the sequence, resulting in the model output.</p>
<p>Interestingly, the KQV layers are located within a ResNet submodule, and the transformer network includes numerous attention modules. Li et al. (<xref ref-type="bibr" rid="ref121">121</xref>) have extended this paradigm to image analysis and accurately classified AD patients by training the attention submodule to identify brain features that maximize patient disease status.</p>
<p>It is worth noting that large-language models can be trained using only ternary values (&#x2212;1, 0, and 1) instead of 16-bit floating-point numbers, according to Ma et al. (<xref ref-type="bibr" rid="ref122">122</xref>). This technique is more precise than converting weights to ternary values after training, resulting in smaller memory footprints and faster response times. This approach is expected to yield better results when applied to healthcare-related solutions.</p>
</sec>
<sec id="sec20">
<label>3.2.4</label>
<title>Graph neural networks</title>
<p>Graph Neural Networks (GNNs) consider the interactions between modeled entities and the topological information in those interactions. They can be used to predict the state of a node based on an input that specifies the state of other nodes and the state of unknown edges. They can also be used to generate new subgraphs in a given network. GNNs can be combined with autoencoder architectures, where the input is first encoded into a low-dimensional graph that performs predictions and then decoded back into the original input representation. Additionally, GNNs can implement the attention mechanism from transformer architectures by learning a weight vector that selects which neighbors to consider during prediction. Truong-Quoc et al. (<xref ref-type="bibr" rid="ref123">123</xref>) used the contextual information used by GNNs to infer 3D spatial DNA conformation despite a scarcity of experimental DNA spatial configuration data. In the field of biological clocks, In&#x00E1;cio et al. (<xref ref-type="bibr" rid="ref124">124</xref>) modeled cardiac shape as a 3D mesh and trained a GNN to infer the biological age from the mesh dynamic motion. Furthermore, Gao et al. (<xref ref-type="bibr" rid="ref125">125</xref>) predicted brain age from MRI data, which was used to compute connectivity maps input into a DNN combining GNN with a transformer.</p>
<p>Remarkably, exploring interactions among multiple entities simultaneously has spurred the creation of novel neural network architectures (<xref ref-type="bibr" rid="ref126">126</xref>). In the realm of Topological Neural Networks, the focus lies in capturing higher-order relational structures, exemplified by edges with multiple sources. These structures unveil intricate interactions within the training data. Considering that biological processes hinge on reactions involving multiple substrates and catalysts (<xref ref-type="bibr" rid="ref67">67</xref>), topological neural networks are poised to provide more effective modeling capabilities for such scenarios.</p>
</sec>
<sec id="sec21">
<label>3.2.5</label>
<title>Ensemble techniques and distributed artificial intelligence</title>
<p>Ensemble techniques allow the creation of comprehensive models by combining multiple application-specific prediction models. For instance, Shah et al. (<xref ref-type="bibr" rid="ref127">127</xref>) could predict biological heart age from MRI by merging the outputs of hundreds of distinct segmentation models to identify cardiac features. Similarly, Huang et al. (<xref ref-type="bibr" rid="ref128">128</xref>) used two parallel networks to extract brain imaging features from MRI and positron-emission tomography (PET), whose outputs were combined to predict patient AD status.</p>
<p>These ensemble techniques offer the opportunity to generate all-inclusive models by integrating multi-modality patient data. However, this integration should respect patient privacy and security and not infringe on intellectual property rights over a dataset. Wassan et al. (<xref ref-type="bibr" rid="ref113">113</xref>), Acosta et al. (<xref ref-type="bibr" rid="ref129">129</xref>), and Wang et al. (<xref ref-type="bibr" rid="ref111">111</xref>) have highlighted these concerns. Federated Learning (FL) approaches have been developed to address these issues. FL formalizes a set of protocols independent collaborators can follow to train a prediction model while keeping their datasets private.</p>
<p>FL generally involves a centralized party coordinating model training by sharing the Deep Neural Network (DNN) weights with each collaborator. The collaborators update the weights based on the data locally available (<xref ref-type="bibr" rid="ref56">56</xref>, <xref ref-type="bibr" rid="ref129">129</xref>, <xref ref-type="bibr" rid="ref130">130</xref>). Decentralized FL approaches, such as the one proposed by Kalra et al. (<xref ref-type="bibr" rid="ref131">131</xref>), are also possible. Collaborators communicate a proxy model trained with individual private models in this protocol. The training is designed in a way that the predictions of the proxy model are also considered in the private model training (<xref ref-type="bibr" rid="ref132">132</xref>). The private model updates its weights considering the datasets used during proxy model training. Conversely, the proxy model is refined with new information from the local dataset before being sent to the next collaborator.</p>
</sec>
</sec>
<sec id="sec22">
<label>3.3</label>
<title>Modeling nonlinear phenomena</title>
<p>The surge in nonlinear methods for GRN reconstruction is driven by the abundance of available data (<xref ref-type="bibr" rid="ref19">19</xref>, <xref ref-type="bibr" rid="ref30">30</xref>, <xref ref-type="bibr" rid="ref68">68</xref>, <xref ref-type="bibr" rid="ref133 ref134 ref135">133&#x2013;135</xref>), offers a versatile framework. This nonlinear description enables the synthesis of diverse findings in the literature (<xref ref-type="bibr" rid="ref32">32</xref>, <xref ref-type="bibr" rid="ref51">51</xref>, <xref ref-type="bibr" rid="ref57">57</xref>, <xref ref-type="bibr" rid="ref58">58</xref>). Notably, studies across tissues or the same tissue under various treatments occupy different regions in the gene expression state space, forming trajectories (<xref ref-type="bibr" rid="ref133">133</xref>). These trajectories reflect the response to specific treatments and are approximated by linear planes with distinct inclinations tangent to the epigenetic landscape (<xref ref-type="fig" rid="fig5">Figures 5A</xref>,<xref ref-type="fig" rid="fig5">B</xref>). The overall epigenetic landscape can be locally described through a combination of linear approaches, even if results from different experiments yield planes with different inclinations, also represented by correlation GRNs with a distinct set of edges (<xref ref-type="fig" rid="fig5">Figures 5A</xref>,<xref ref-type="fig" rid="fig5">C</xref>,<xref ref-type="fig" rid="fig5">D</xref>).</p>
<fig position="float" id="fig5">
<label>Figure 5</label>
<caption>
<p>Integrating multiple linear approaches to unravel a nonlinear landscape. <bold>(A)</bold> The epigenetic landscape as a function of the expression level of two master regulator genes X and Y. Experiments analyzing linear relationships in genes responding to a particular treatment unravel the underlying landscape by finding linear approximations, illustrated by planes tangent to the landscape at the indicated points. Each point is associated with an experimental model, presenting specific X and Y basal expression values. <bold>(B)</bold> At points B1 and B2, genes A and B, downstream to X and Y, show distinct correlations. <bold>(C)</bold> Correlation matrices depicting pairwise correlations for genes downstream of X and Y at points C1, D1; C2, D2; and C3, D3. The different correlation matrices associated with planes with distinct inclinations at each point approximate the landscape curvature. <bold>(D)</bold> Correlation networks corresponding to each correlation matrix in C. As pairwise gene correlations assume different values depending on the expression levels of X and Y, i.e., depending on the position in the landscape, the gene correlation network contains a distinct set of edges.</p>
</caption>
<graphic xlink:href="fams-10-1380996-g005.tif"/>
</fig>
<p>Within a single experiment, linear correlations determine the plane inclination at each state, leading to varied regression coefficients or pairwise correlations across different experiments in distinct conditions. Alternatively, the linear regression coefficient between two molecules can be viewed as dependent on the expression of a specific regulator (<xref ref-type="bibr" rid="ref67">67</xref>). Modulating the regulator expression mirrors sensitivity analysis (<xref ref-type="bibr" rid="ref69">69</xref>), visualized by rearranging the epigenetic landscape, akin to Li et al.&#x2019;s (<xref ref-type="bibr" rid="ref70">70</xref>) landscape analysis for distinct SIRT levels. Existing GRN descriptions can be perceived as a comprehensive piece-wise linear model, accommodating batch effects from measurement collection in diverse laboratories.</p>
<p>Considering the cellular ability to dedifferentiate and transdifferentiate, allowing access to the entire epigenetic landscape even in differentiated cells, Shin and Cho (<xref ref-type="bibr" rid="ref19">19</xref>) emphasize the importance of gathering data on intermediate states to enhance the accuracy of GRN evolution models. Echoing this, our group&#x2019;s prior work illustrates how dynamic datasets contribute to a more thorough comprehension of cell fate decisions (<xref ref-type="bibr" rid="ref51">51</xref>, <xref ref-type="bibr" rid="ref52">52</xref>).</p>
</sec>
<sec id="sec23">
<label>3.4</label>
<title>Higher-order dynamic representations</title>
<p>ODE approaches commonly express a species&#x2019; rate of change as a function of its current state, a concept that can be extended to higher-order systems where acceleration is tied to the system&#x2019;s state. Schwanh&#x00E4;usser et al. (<xref ref-type="bibr" rid="ref14">14</xref>) emphasized the importance of considering gene expression levels proportional to the protein translation rate for accurate protein level inference from transcriptome data. In a study on monocyte differentiation during heart failure, Ni et al. (<xref ref-type="bibr" rid="ref136">136</xref>) discovered the Cd72 gene&#x2019;s expression derivative peaked when its transcriptional regulator RelA was maximally expressed.</p>
<p>RNA velocity, distinguishing spliced and unspliced RNA molecules, is increasingly employed in single-cell transcriptome analysis (<xref ref-type="bibr" rid="ref15">15</xref>, <xref ref-type="bibr" rid="ref137">137</xref>), using ratios to infer gene expression rates. However, the appropriate dynamic order for modeling biological systems remains to be determined due to the complexity of signaling pathways and regulatory layers (<xref ref-type="bibr" rid="ref16">16</xref>, <xref ref-type="bibr" rid="ref17">17</xref>). Speculating on effective comparisons, it might be advantageous to model histone mark modification and transcription factor binding as a second-order dynamic system (<xref ref-type="bibr" rid="ref138">138</xref>). Alternatively, comparing enzyme levels instead of metabolite production requires accurate measurement of all implicated regulation layers, acknowledging the challenge posed by the inherent variability in protein levels despite consistent RNA molecule counts.</p>
<p>Theoretically, second-order systems offer robustness, allowing a broader range of dynamic processes. Networks with self-regulatory loops described by first-order systems can only exhibit exponential decay or unbounded increase (<xref ref-type="fig" rid="fig4">Figure 4D</xref>). In contrast, equivalent networks involving proteins regulating their genes can display these two behaviors with the addition of oscillations (<xref ref-type="bibr" rid="ref139">139</xref>). <xref ref-type="fig" rid="fig6">Figure 6A</xref> illustrates an extension of the example GRN in <xref ref-type="fig" rid="fig4">Figure 4A</xref>, where the expression of the target gene T considers both freshly transcribed unspliced mRNA (T<sub>U</sub>) and mature spliced mRNA levels (T<sub>S</sub>). T<sub>S</sub> is increasingly synthesized with higher T<sub>U</sub> levels and is regulated by temporal degradation. The equation governing this dynamic is depicted in <xref ref-type="fig" rid="fig6">Figure 6B</xref>, which, after some algebraic manipulation, is equivalent to the second-order ODE shown in <xref ref-type="fig" rid="fig6">Figure 6C</xref>.</p>
<fig position="float" id="fig6">
<label>Figure 6</label>
<caption>
<p>Higher-order differential equations. <bold>(A)</bold> Gene regulatory network (GRN) portraying a dynamic transcription model, where a target gene T produces unspliced mRNA fragments (T<sub>U</sub>), which are spliced into mature mRNA (T<sub>S</sub>). The regulator R determines the stable expression level of its target T<sub>S</sub>. <bold>(B)</bold> Differential equations describing T<sub>U</sub> production for mature mRNA T<sub>S</sub> and R transcription levels. T<sub>s</sub> is produced proportionally to T<sub>U</sub> abundance and decays over time. <bold>(C)</bold> A second-order differential equation equivalent to the equations system shown in panel <bold>(B)</bold>. <bold>(D)</bold> Temporal T<sub>s</sub> expression for the same R values depicted in <xref ref-type="fig" rid="fig3">Figure 3D</xref>. The dynamics are shown for distinct initial T<sub>s</sub> expression values T<sub>s</sub>0. In regime, T<sub>s</sub> transcription levels tend to R, which are reached after a transient oscillatory activity. Oscillation frequency is also dependent on R expression levels.</p>
</caption>
<graphic xlink:href="fams-10-1380996-g006.tif"/>
</fig>
<p>Similarly to <xref ref-type="fig" rid="fig3">Figure 3D</xref>, which demonstrates the system convergence to the attractor T&#x2009;=&#x2009;R, <xref ref-type="fig" rid="fig6">Figure 6D</xref> shows that the second-order ODE also leads to equivalent attractors in the long term. Nonetheless, this state is achieved only after oscillatory behavior. Interestingly, R values determine the system&#x2019;s final state and the oscillation frequency.</p>
<p>Given that enough samples are available, higher-order modeling has the potential to implicitly consider layers of regulation that are not present in the training data. This process is analogous to the equation derivation in <xref ref-type="fig" rid="fig6">Figure 6C</xref>, which does not include T<sub>U</sub>. Nonetheless, T<sub>S</sub> expression numeric simulations still require an initial value for its derivative, equivalent to assuming T<sub>U</sub> initial values. This shortcoming can be overcome by performing multiple simulations with distinct initial values for unknown variables at the cost of additional computing resources or by estimating these values based on data from a similar cell type (e.g., using the atlases depicted in <xref ref-type="table" rid="tab2">Table 2</xref> as potential sources). Altogether, the versatility of higher-order dynamic models has great potential to increase the predictive power for GRN inference.</p>
</sec>
</sec>
<sec sec-type="conclusions" id="sec24">
<label>4</label>
<title>Conclusion</title>
<p>The study of aging has achieved notable advancements, unveiling shared processes and biomarkers across species. Despite this progress, a comprehensive understanding of the interconnectedness of age-related molecular processes still needs to be discovered. ARDs likely result from a combination of these processes, inducing pathophysiological shifts. Exploring strategies mapping the epigenetic landscape of age-related cellular abnormalities has become more refined with recent breakthroughs in single-cell data acquisition and integrating diverse biochemical regulation layers. These advancements present an unprecedented opportunity to unravel the molecular intricacies of aging, offering the potential for therapies employing dynamic systems concepts to redirect pathological trajectories toward a healthier course.</p>
<p>Identifying shared aging mechanisms across multiple animal species and tissue types has unveiled that interactions among genes, proteins, metabolites, and epigenetic marks form a complex, intricate network spanning processes from molecular to organismal levels. Hence, these processes are characterized by many parameters, requiring an equally extensive amount of data to distinguish GRNs that generalize results from multiple independent experiments.</p>
<p>Recent advances in experimental techniques and data analysis poise the aging field to disruptive breakthroughs in identifying age-associated trajectories and the delay in the onset of age-related pathologies. The solidification of the FAIR principles (findability, accessibility, interoperability, and reusability) in data sharing has paved the way to integrating comprehensive datasets that are impossible through multiple isolated studies. These include the identification of cellular heterogeneity and dynamics in aging processes through integrating multi-layered networks, including GRNs and PPI networks, single-cell sequencing experiments, and spatial molecular profiling assays. We believe that the knowledge gained from systems pharmacology will contribute to the widespread use of computational models in healthcare. These models will help in identifying and prioritizing potential interventions to modify the pathways related to aging in a personalized way. This approach will take into account an individual&#x2019;s genetic background, lifestyle factors, and environmental exposures and will be crucial in developing effective anti-aging therapies.</p>
</sec>
<sec sec-type="author-contributions" id="sec25">
<title>Author contributions</title>
<p>JF: Conceptualization, Funding acquisition, Resources, Visualization, Writing &#x2013; original draft, Writing &#x2013; review &#x0026; editing. OB: Conceptualization, Funding acquisition, Resources, Supervision, Writing &#x2013; original draft, Writing &#x2013; review &#x0026; editing.</p>
</sec>
</body>
<back>
<sec sec-type="funding-information" id="sec26">
<title>Funding</title>
<p>The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This work was supported by grants from the INSERM AGEMED program and ANR-19-CE13-0017&#x2009;S-ENCODE to OB. JF was supported by the &#x201C;MATURATION Jeunes chercheurs de la Facult&#x00E9; de Sant&#x00E9; Paris Est Cr&#x00E9;teil&#x201D; grant.</p>
</sec>
<ack>
<p>We are grateful to the members of INSERM-U955 for their support, critical reading of the manuscript, and helpful discussions.</p>
</ack>
<sec sec-type="COI-statement" id="sec27">
<title>Conflict of interest</title>
<p>The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.</p>
</sec>
<sec sec-type="disclaimer" id="sec28">
<title>Publisher&#x2019;s note</title>
<p>All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.</p>
</sec>
<ref-list>
<title>References</title>
<ref id="ref1">
<label>1.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Jones</surname> <given-names>OR</given-names></name> <name><surname>Scheuerlein</surname> <given-names>A</given-names></name> <name><surname>Salguero-G&#x00F3;mez</surname> <given-names>R</given-names></name> <name><surname>Camarda</surname> <given-names>CG</given-names></name> <name><surname>Schaible</surname> <given-names>R</given-names></name> <name><surname>Casper</surname> <given-names>BB</given-names></name> <etal/></person-group>. <article-title>Diversity of ageing across the tree of life</article-title>. <source>Nature</source>. (<year>2014</year>) <volume>505</volume>:<fpage>169</fpage>&#x2013;<lpage>73</lpage>. doi: <pub-id pub-id-type="doi">10.1038/nature12789</pub-id>, PMID: <pub-id pub-id-type="pmid">24317695</pub-id></citation>
</ref>
<ref id="ref2">
<label>2.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Santiago</surname> <given-names>E</given-names></name> <name><surname>Moreno</surname> <given-names>DF</given-names></name> <name><surname>Acar</surname> <given-names>M</given-names></name></person-group>. <article-title>Modeling aging and its impact on cellular function and organismal behavior</article-title>. <source>Exp Gerontol</source>. (<year>2021</year>) <volume>155</volume>:<fpage>111577</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.exger.2021.111577</pub-id>, PMID: <pub-id pub-id-type="pmid">34582969</pub-id></citation>
</ref>
<ref id="ref3">
<label>3.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Gompertz</surname> <given-names>B</given-names></name>
</person-group>. <article-title>XXIV. On the nature of the function expressive of the law of human mortality, and on a new mode of determining the value of life contingencies. In a letter to Francis baily, Esq. FRS &#x0026;c</article-title>. <source>Philos Trans R Soc Lond</source>. (<year>1825</year>) <volume>115</volume>:<fpage>513</fpage>&#x2013;<lpage>83</lpage>. doi: <pub-id pub-id-type="doi">10.1098/rstl.1825.0026</pub-id></citation>
</ref>
<ref id="ref4">
<label>4.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Kappeler</surname> <given-names>L</given-names></name> <name><surname>Filho</surname> <given-names>CDM</given-names></name> <name><surname>Dupont</surname> <given-names>J</given-names></name> <name><surname>Leneuve</surname> <given-names>P</given-names></name> <name><surname>Cervera</surname> <given-names>P</given-names></name> <name><surname>P&#x00E9;rin</surname> <given-names>L</given-names></name> <etal/></person-group>. <article-title>Brain IGF-1 receptors control mammalian growth and lifespan through a neuroendocrine mechanism</article-title>. <source>PLoS Biol</source>. (<year>2008</year>) <volume>6</volume>:<fpage>e254</fpage>. doi: <pub-id pub-id-type="doi">10.1371/journal.pbio.0060254</pub-id>, PMID: <pub-id pub-id-type="pmid">18959478</pub-id></citation>
</ref>
<ref id="ref5">
<label>5.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Kenyon</surname> <given-names>C</given-names></name> <name><surname>Chang</surname> <given-names>J</given-names></name> <name><surname>Gensch</surname> <given-names>E</given-names></name> <name><surname>Rudner</surname> <given-names>A</given-names></name> <name><surname>Tabtiang</surname> <given-names>R</given-names></name></person-group>. <article-title>A <italic>C. elegans</italic> mutant that lives twice as long as wild type</article-title>. <source>Nature</source>. (<year>1993</year>) <volume>366</volume>:<fpage>461</fpage>&#x2013;<lpage>4</lpage>. doi: <pub-id pub-id-type="doi">10.1038/366461a0</pub-id>, PMID: <pub-id pub-id-type="pmid">8247153</pub-id></citation>
</ref>
<ref id="ref6">
<label>6.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Kenyon</surname> <given-names>CJ</given-names></name>
</person-group>. <article-title>The genetics of ageing</article-title>. <source>Nature</source>. (<year>2010</year>) <volume>464</volume>:<fpage>504</fpage>&#x2013;<lpage>12</lpage>. doi: <pub-id pub-id-type="doi">10.1038/nature08980</pub-id></citation>
</ref>
<ref id="ref7">
<label>7.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Kenyon</surname> <given-names>C</given-names></name>
</person-group>. <article-title>The first long-lived mutants: discovery of the insulin/IGF-1 pathway for ageing</article-title>. <source>Phil Trans R Soc B: Biol Sci</source>. (<year>2011</year>) <volume>366</volume>:<fpage>9</fpage>&#x2013;<lpage>16</lpage>. doi: <pub-id pub-id-type="doi">10.1098/rstb.2010.0276</pub-id>, PMID: <pub-id pub-id-type="pmid">21115525</pub-id></citation>
</ref>
<ref id="ref8">
<label>8.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Mc Auley</surname> <given-names>MT</given-names></name> <name><surname>Guimera</surname> <given-names>AM</given-names></name> <name><surname>Hodgson</surname> <given-names>D</given-names></name> <name><surname>Mcdonald</surname> <given-names>N</given-names></name> <name><surname>Mooney</surname> <given-names>KM</given-names></name> <name><surname>Morgan</surname> <given-names>AE</given-names></name> <etal/></person-group>. <article-title>Modelling the molecular mechanisms of aging</article-title>. <source>Biosci Rep</source>. (<year>2017</year>) <volume>37</volume>:<fpage>BSR20160177</fpage>. doi: <pub-id pub-id-type="doi">10.1042/BSR20160177</pub-id>, PMID: <pub-id pub-id-type="pmid">28096317</pub-id></citation>
</ref>
<ref id="ref9">
<label>9.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>McCormick</surname> <given-names>MA</given-names></name> <name><surname>Promislow</surname> <given-names>DE</given-names></name></person-group>. <article-title>Recent advances in the systems biology of aging</article-title>. <source>Antioxid Redox Signal</source>. (<year>2018</year>) <volume>29</volume>:<fpage>973</fpage>&#x2013;<lpage>84</lpage>. doi: <pub-id pub-id-type="doi">10.1089/ars.2017.7367</pub-id>, PMID: <pub-id pub-id-type="pmid">29020802</pub-id></citation>
</ref>
<ref id="ref10">
<label>10.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Melzer</surname> <given-names>D</given-names></name> <name><surname>Pilling</surname> <given-names>LC</given-names></name> <name><surname>Ferrucci</surname> <given-names>L</given-names></name></person-group>. <article-title>The genetics of human ageing</article-title>. <source>Nat Rev Genet</source>. (<year>2020</year>) <volume>21</volume>:<fpage>88</fpage>&#x2013;<lpage>101</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41576-019-0183-6</pub-id>, PMID: <pub-id pub-id-type="pmid">31690828</pub-id></citation>
</ref>
<ref id="ref11">
<label>11.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Tacutu</surname> <given-names>R</given-names></name> <name><surname>Thornton</surname> <given-names>D</given-names></name> <name><surname>Johnson</surname> <given-names>E</given-names></name> <name><surname>Budovsky</surname> <given-names>A</given-names></name> <name><surname>Barardo</surname> <given-names>D</given-names></name> <name><surname>Craig</surname> <given-names>T</given-names></name> <etal/></person-group>. <article-title>Human ageing genomic resources: new and updated databases</article-title>. <source>Nucleic Acids Res</source>. (<year>2018</year>) <volume>46</volume>:<fpage>D1083</fpage>&#x2013;<lpage>90</lpage>. doi: <pub-id pub-id-type="doi">10.1093/nar/gkx1042</pub-id>, PMID: <pub-id pub-id-type="pmid">29121237</pub-id></citation>
</ref>
<ref id="ref12">
<label>12.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>L&#x00F3;pez-Ot&#x00ED;n</surname> <given-names>C</given-names></name> <name><surname>Blasco</surname> <given-names>MA</given-names></name> <name><surname>Partridge</surname> <given-names>L</given-names></name> <name><surname>Serrano</surname> <given-names>M</given-names></name> <name><surname>Kroemer</surname> <given-names>G</given-names></name></person-group>. <article-title>The hallmarks of aging</article-title>. <source>Cell</source>. (<year>2013</year>) <volume>153</volume>:<fpage>1194</fpage>&#x2013;<lpage>217</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.cell.2013.05.039</pub-id>, PMID: <pub-id pub-id-type="pmid">23746838</pub-id></citation>
</ref>
<ref id="ref13">
<label>13.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>L&#x00F3;pez-Ot&#x00ED;n</surname> <given-names>C</given-names></name> <name><surname>Blasco</surname> <given-names>MA</given-names></name> <name><surname>Partridge</surname> <given-names>L</given-names></name> <name><surname>Serrano</surname> <given-names>M</given-names></name> <name><surname>Kroemer</surname> <given-names>G</given-names></name></person-group>. <article-title>Hallmarks of aging: an expanding universe</article-title>. <source>Cell</source>. (<year>2023</year>) <volume>186</volume>:<fpage>243</fpage>&#x2013;<lpage>78</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.cell.2022.11.001</pub-id>, PMID: <pub-id pub-id-type="pmid">36599349</pub-id></citation>
</ref>
<ref id="ref14">
<label>14.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Schwanh&#x00E4;usser</surname> <given-names>B</given-names></name> <name><surname>Busse</surname> <given-names>D</given-names></name> <name><surname>Li</surname> <given-names>N</given-names></name> <name><surname>Dittmar</surname> <given-names>G</given-names></name> <name><surname>Schuchhardt</surname> <given-names>J</given-names></name> <name><surname>Wolf</surname> <given-names>J</given-names></name> <etal/></person-group>. <article-title>Global quantification of mammalian gene expression control</article-title>. <source>Nature</source>. (<year>2011</year>) <volume>473</volume>:<fpage>337</fpage>&#x2013;<lpage>42</lpage>. doi: <pub-id pub-id-type="doi">10.1038/nature10098</pub-id></citation>
</ref>
<ref id="ref15">
<label>15.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>La Manno</surname> <given-names>G</given-names></name> <name><surname>Soldatov</surname> <given-names>R</given-names></name> <name><surname>Zeisel</surname> <given-names>A</given-names></name> <name><surname>Braun</surname> <given-names>E</given-names></name> <name><surname>Hochgerner</surname> <given-names>H</given-names></name> <name><surname>Petukhov</surname> <given-names>V</given-names></name> <etal/></person-group>. <article-title>RNA velocity of single cells</article-title>. <source>Nature</source>. (<year>2018</year>) <volume>560</volume>:<fpage>494</fpage>&#x2013;<lpage>8</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41586-018-0414-6</pub-id>, PMID: <pub-id pub-id-type="pmid">30089906</pub-id></citation>
</ref>
<ref id="ref16">
<label>16.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Hoffmann</surname> <given-names>M</given-names></name> <name><surname>Fr&#x00F6;hner</surname> <given-names>C</given-names></name> <name><surname>No&#x00E9;</surname> <given-names>F</given-names></name></person-group>. <article-title>Reactive SINDy: discovering governing reactions from concentration data</article-title>. <source>J Chem Phys</source>. (<year>2019</year>) <volume>150</volume>:<fpage>025101</fpage>. doi: <pub-id pub-id-type="doi">10.1063/1.5066099</pub-id>, PMID: <pub-id pub-id-type="pmid">30646700</pub-id></citation>
</ref>
<ref id="ref17">
<label>17.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Badia-i-Mompel</surname> <given-names>P</given-names></name> <name><surname>Wessels</surname> <given-names>L</given-names></name> <name><surname>M&#x00FC;ller-Dott</surname> <given-names>S</given-names></name> <name><surname>Trimbour</surname> <given-names>R</given-names></name> <name><surname>Ramirez Flores</surname> <given-names>RO</given-names></name> <name><surname>Argelaguet</surname> <given-names>R</given-names></name> <etal/></person-group>. <article-title>Gene regulatory network inference in the era of single-cell multi-omics</article-title>. <source>Nat Rev Genet</source>. (<year>2023</year>) <volume>24</volume>:<fpage>739</fpage>&#x2013;<lpage>54</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41576-023-00618-5</pub-id>, PMID: <pub-id pub-id-type="pmid">37365273</pub-id></citation>
</ref>
<ref id="ref18">
<label>18.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Chuang</surname> <given-names>HY</given-names></name> <name><surname>Hofree</surname> <given-names>M</given-names></name> <name><surname>Ideker</surname> <given-names>T</given-names></name></person-group>. <article-title>A decade of systems biology</article-title>. <source>Annu Rev Cell Dev Biol</source>. (<year>2010</year>) <volume>26</volume>:<fpage>721</fpage>&#x2013;<lpage>44</lpage>. doi: <pub-id pub-id-type="doi">10.1146/annurev-cellbio-100109-104122</pub-id>, PMID: <pub-id pub-id-type="pmid">20604711</pub-id></citation>
</ref>
<ref id="ref19">
<label>19.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Shin</surname> <given-names>D</given-names></name> <name><surname>Cho</surname> <given-names>KH</given-names></name></person-group>. <article-title>Critical transition and reversion of tumorigenesis</article-title>. <source>Exp Mol Med</source>. (<year>2023</year>) <volume>55</volume>:<fpage>692</fpage>&#x2013;<lpage>705</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s12276-023-00969-3</pub-id>, PMID: <pub-id pub-id-type="pmid">37009794</pub-id></citation>
</ref>
<ref id="ref20">
<label>20.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Wysocka</surname> <given-names>M</given-names></name> <name><surname>Wysocki</surname> <given-names>O</given-names></name> <name><surname>Zufferey</surname> <given-names>M</given-names></name> <name><surname>Landers</surname> <given-names>D</given-names></name> <name><surname>Freitas</surname> <given-names>A</given-names></name></person-group>. <article-title>A systematic review of biologically-informed deep learning models for cancer: fundamental trends for encoding and interpreting oncology data</article-title>. <source>BMC Bioinf</source>. (<year>2023</year>) <volume>24</volume>:<fpage>198</fpage>&#x2013;<lpage>31</lpage>. doi: <pub-id pub-id-type="doi">10.1186/s12859-023-05262-8</pub-id>, PMID: <pub-id pub-id-type="pmid">37189058</pub-id></citation>
</ref>
<ref id="ref21">
<label>21.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Kogan</surname> <given-names>V</given-names></name> <name><surname>Molodtsov</surname> <given-names>I</given-names></name> <name><surname>Menshikov</surname> <given-names>LI</given-names></name> <name><surname>Reis</surname> <given-names>RJS</given-names></name> <name><surname>Fedichev</surname> <given-names>P</given-names></name></person-group>. <article-title>Stability analysis of a model gene network links aging, stress resistance and negligible senescence</article-title>. <source>Sci Rep</source>. (<year>2015</year>) <volume>5</volume>:<fpage>13589</fpage>. doi: <pub-id pub-id-type="doi">10.1038/srep13589</pub-id>, PMID: <pub-id pub-id-type="pmid">26316217</pub-id></citation>
</ref>
<ref id="ref22">
<label>22.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Galvis</surname> <given-names>D</given-names></name> <name><surname>Walsh</surname> <given-names>D</given-names></name> <name><surname>Harries</surname> <given-names>LW</given-names></name> <name><surname>Latorre</surname> <given-names>E</given-names></name> <name><surname>Rankin</surname> <given-names>J</given-names></name></person-group>. <article-title>A dynamical systems model for the measurement of cellular senescence</article-title>. <source>J R Soc Interface</source>. (<year>2019</year>) <volume>16</volume>:<fpage>20190311</fpage>. doi: <pub-id pub-id-type="doi">10.1098/rsif.2019.0311</pub-id>, PMID: <pub-id pub-id-type="pmid">31594522</pub-id></citation>
</ref>
<ref id="ref23">
<label>23.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Choo</surname> <given-names>SM</given-names></name> <name><surname>Ban</surname> <given-names>B</given-names></name> <name><surname>Joo</surname> <given-names>JI</given-names></name> <name><surname>Cho</surname> <given-names>KH</given-names></name></person-group>. <article-title>The phenotype control kernel of a biomolecular regulatory network</article-title>. <source>BMC Syst Biol</source>. (<year>2018</year>) <volume>12</volume>:<fpage>49</fpage>&#x2013;<lpage>15</lpage>. doi: <pub-id pub-id-type="doi">10.1186/s12918-018-0576-8</pub-id>, PMID: <pub-id pub-id-type="pmid">29622038</pub-id></citation>
</ref>
<ref id="ref24">
<label>24.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Schwab</surname> <given-names>JD</given-names></name> <name><surname>Siegle</surname> <given-names>L</given-names></name> <name><surname>K&#x00FC;hlwein</surname> <given-names>SD</given-names></name> <name><surname>K&#x00FC;hl</surname> <given-names>M</given-names></name> <name><surname>Kestler</surname> <given-names>HA</given-names></name></person-group>. <article-title>Stability of signaling pathways during aging&#x2014;a Boolean network approach</article-title>. <source>Biology</source>. (<year>2017</year>) <volume>6</volume>:<fpage>46</fpage>. doi: <pub-id pub-id-type="doi">10.3390/biology6040046</pub-id>, PMID: <pub-id pub-id-type="pmid">29258225</pub-id></citation>
</ref>
<ref id="ref25">
<label>25.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Schwab</surname> <given-names>JD</given-names></name> <name><surname>Ikonomi</surname> <given-names>N</given-names></name> <name><surname>Werle</surname> <given-names>SD</given-names></name> <name><surname>Weidner</surname> <given-names>FM</given-names></name> <name><surname>Geiger</surname> <given-names>H</given-names></name> <name><surname>Kestler</surname> <given-names>HA</given-names></name></person-group>. <article-title>Reconstructing Boolean network ensembles from single-cell data for unraveling dynamics in the aging of human hematopoietic stem cells</article-title>. <source>Comput Struct Biotechnol J</source>. (<year>2021</year>) <volume>19</volume>:<fpage>5321</fpage>&#x2013;<lpage>32</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.csbj.2021.09.012</pub-id>, PMID: <pub-id pub-id-type="pmid">34630946</pub-id></citation>
</ref>
<ref id="ref26">
<label>26.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Al Qazlan</surname> <given-names>TA</given-names></name> <name><surname>Hamdi-Cherif</surname> <given-names>A</given-names></name> <name><surname>Kara-Mohamed</surname> <given-names>C</given-names></name></person-group>. <article-title>State of the art of fuzzy methods for gene regulatory networks inference</article-title>. <source>Sci World J</source>. (<year>2015</year>) <volume>2015</volume>:<fpage>1</fpage>&#x2013;<lpage>11</lpage>. doi: <pub-id pub-id-type="doi">10.1155/2015/148010</pub-id>, PMID: <pub-id pub-id-type="pmid">25879048</pub-id></citation>
</ref>
<ref id="ref27">
<label>27.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Raza</surname> <given-names>K</given-names></name>
</person-group>. <article-title>Fuzzy logic-based approaches for gene regulatory network inference</article-title>. <source>Artif Intell Med</source>. (<year>2019</year>) <volume>97</volume>:<fpage>189</fpage>&#x2013;<lpage>203</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.artmed.2018.12.004</pub-id>, PMID: <pub-id pub-id-type="pmid">30573378</pub-id></citation>
</ref>
<ref id="ref28">
<label>28.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Aldridge</surname> <given-names>BB</given-names></name> <name><surname>Saez-Rodriguez</surname> <given-names>J</given-names></name> <name><surname>Muhlich</surname> <given-names>JL</given-names></name> <name><surname>Sorger</surname> <given-names>PK</given-names></name> <name><surname>Lauffenburger</surname> <given-names>DA</given-names></name></person-group>. <article-title>Fuzzy logic analysis of kinase pathway crosstalk in TNF/EGF/insulin-induced signaling</article-title>. <source>PLoS Comput Biol</source>. (<year>2009</year>) <volume>5</volume>:<fpage>e1000340</fpage>. doi: <pub-id pub-id-type="doi">10.1371/journal.pcbi.1000340</pub-id>, PMID: <pub-id pub-id-type="pmid">19343194</pub-id></citation>
</ref>
<ref id="ref29">
<label>29.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Poblete</surname> <given-names>CM</given-names></name> <name><surname>Parra</surname> <given-names>FV</given-names></name> <name><surname>Gomez</surname> <given-names>JB</given-names></name> <name><surname>Saldias</surname> <given-names>MC</given-names></name> <name><surname>Garrido</surname> <given-names>SS</given-names></name> <name><surname>Vargas</surname> <given-names>HM</given-names></name></person-group>. <article-title>Fuzzy logic in genetic regulatory network models</article-title>. <source>Int. J. Comput. Commun. Control</source>. (<year>2009</year>) <volume>4</volume>:<fpage>363</fpage>&#x2013;<lpage>73</lpage>. doi: <pub-id pub-id-type="doi">10.15837/ijccc.2009.4.2453</pub-id></citation>
</ref>
<ref id="ref30">
<label>30.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Linde</surname> <given-names>J</given-names></name> <name><surname>Schulze</surname> <given-names>S</given-names></name> <name><surname>Henkel</surname> <given-names>SG</given-names></name> <name><surname>Guthke</surname> <given-names>R</given-names></name></person-group>. <article-title>Data-and knowledge-based modeling of gene regulatory networks: an update</article-title>. <source>EXCLI J</source>. (<year>2015</year>) <volume>14</volume>:<fpage>346</fpage>. doi: <pub-id pub-id-type="doi">10.17179/excli2015-168</pub-id></citation>
</ref>
<ref id="ref31">
<label>31.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Hackett</surname> <given-names>SR</given-names></name> <name><surname>Baltz</surname> <given-names>EA</given-names></name> <name><surname>Coram</surname> <given-names>M</given-names></name> <name><surname>Wranik</surname> <given-names>BJ</given-names></name> <name><surname>Kim</surname> <given-names>G</given-names></name> <name><surname>Baker</surname> <given-names>A</given-names></name> <etal/></person-group>. <article-title>Learning causal networks using inducible transcription factors and transcriptome-wide time series</article-title>. <source>Mol Syst Biol</source>. (<year>2020</year>) <volume>16</volume>:<fpage>e9174</fpage>. doi: <pub-id pub-id-type="doi">10.15252/msb.20199174</pub-id>, PMID: <pub-id pub-id-type="pmid">32181581</pub-id></citation>
</ref>
<ref id="ref32">
<label>32.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Li</surname> <given-names>CW</given-names></name> <name><surname>Wang</surname> <given-names>WH</given-names></name> <name><surname>Chen</surname> <given-names>BS</given-names></name></person-group>. <article-title>Investigating the specific core genetic-and-epigenetic networks of cellular mechanisms involved in human aging in peripheral blood mononuclear cells</article-title>. <source>Oncotarget</source>. (<year>2016</year>) <volume>7</volume>:<fpage>8556</fpage>&#x2013;<lpage>79</lpage>. doi: <pub-id pub-id-type="doi">10.18632/oncotarget.7388</pub-id>, PMID: <pub-id pub-id-type="pmid">26895224</pub-id></citation>
</ref>
<ref id="ref33">
<label>33.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Zhang</surname> <given-names>XM</given-names></name> <name><surname>Liang</surname> <given-names>L</given-names></name> <name><surname>Liu</surname> <given-names>L</given-names></name> <name><surname>Tang</surname> <given-names>MJ</given-names></name></person-group>. <article-title>Graph neural networks and their current applications in bioinformatics</article-title>. <source>Front Genet</source>. (<year>2021</year>) <volume>12</volume>:<fpage>690049</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fgene.2021.690049</pub-id></citation>
</ref>
<ref id="ref34">
<label>34.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Jordan</surname> <given-names>DJ</given-names></name> <name><surname>Miska</surname> <given-names>EA</given-names></name></person-group>. <article-title>Canalisation and plasticity on the developmental manifold of <italic>Caenorhabditis elegans</italic></article-title>. <source>Mol Syst Biol</source>. (<year>2023</year>) <volume>19</volume>:<fpage>e11835</fpage>. doi: <pub-id pub-id-type="doi">10.15252/msb.202311835</pub-id>, PMID: <pub-id pub-id-type="pmid">37850520</pub-id></citation>
</ref>
<ref id="ref35">
<label>35.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Salameh</surname> <given-names>Y</given-names></name> <name><surname>Bejaoui</surname> <given-names>Y</given-names></name> <name><surname>El Hajj</surname> <given-names>N</given-names></name></person-group>. <article-title>DNA methylation biomarkers in aging and age-related diseases</article-title>. <source>Front Genet</source>. (<year>2020</year>) <volume>11</volume>:<fpage>171</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fgene.2020.00171</pub-id>, PMID: <pub-id pub-id-type="pmid">32211026</pub-id></citation>
</ref>
<ref id="ref36">
<label>36.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Hannum</surname> <given-names>G</given-names></name> <name><surname>Guinney</surname> <given-names>J</given-names></name> <name><surname>Zhao</surname> <given-names>L</given-names></name> <name><surname>Zhang</surname> <given-names>LI</given-names></name> <name><surname>Hughes</surname> <given-names>G</given-names></name> <name><surname>Sadda</surname> <given-names>S</given-names></name> <etal/></person-group>. <article-title>Genome-wide methylation profiles reveal quantitative views of human aging rates</article-title>. <source>Mol Cell</source>. (<year>2013</year>) <volume>49</volume>:<fpage>359</fpage>&#x2013;<lpage>67</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.molcel.2012.10.016</pub-id>, PMID: <pub-id pub-id-type="pmid">23177740</pub-id></citation>
</ref>
<ref id="ref37">
<label>37.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Horvath</surname> <given-names>S</given-names></name>
</person-group>. <article-title>DNA methylation age of human tissues and cell types</article-title>. <source>Genome Biol</source>. (<year>2013</year>) <volume>14</volume>:<fpage>R115</fpage>&#x2013;<lpage>20</lpage>. doi: <pub-id pub-id-type="doi">10.1186/gb-2013-14-10-r115</pub-id>, PMID: <pub-id pub-id-type="pmid">24138928</pub-id></citation>
</ref>
<ref id="ref38">
<label>38.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Levine</surname> <given-names>ME</given-names></name> <name><surname>Lu</surname> <given-names>AT</given-names></name> <name><surname>Quach</surname> <given-names>A</given-names></name> <name><surname>Chen</surname> <given-names>BH</given-names></name> <name><surname>Assimes</surname> <given-names>TL</given-names></name> <name><surname>Bandinelli</surname> <given-names>S</given-names></name> <etal/></person-group>. <article-title>An epigenetic biomarker of aging for lifespan and healthspan</article-title>. <source>Aging (Albany NY)</source>. (<year>2018</year>) <volume>10</volume>:<fpage>573</fpage>&#x2013;<lpage>91</lpage>. doi: <pub-id pub-id-type="doi">10.18632/aging.101414</pub-id>, PMID: <pub-id pub-id-type="pmid">29676998</pub-id></citation>
</ref>
<ref id="ref39">
<label>39.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Belsky</surname> <given-names>DW</given-names></name> <name><surname>Caspi</surname> <given-names>A</given-names></name> <name><surname>Corcoran</surname> <given-names>DL</given-names></name> <name><surname>Sugden</surname> <given-names>K</given-names></name> <name><surname>Poulton</surname> <given-names>R</given-names></name> <name><surname>Arseneault</surname> <given-names>L</given-names></name> <etal/></person-group>. <article-title>DunedinPACE, a DNA methylation biomarker of the pace of aging</article-title>. <source>eLife</source>. (<year>2022</year>) <volume>11</volume>:<fpage>e73420</fpage>. doi: <pub-id pub-id-type="doi">10.7554/eLife.73420</pub-id></citation>
</ref>
<ref id="ref40">
<label>40.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Rappoport</surname> <given-names>N</given-names></name> <name><surname>Shamir</surname> <given-names>R</given-names></name></person-group>. <article-title>Multi-omic and multi-view clustering algorithms: review and cancer benchmark</article-title>. <source>Nucleic Acids Res</source>. (<year>2018</year>) <volume>46</volume>:<fpage>10546</fpage>&#x2013;<lpage>62</lpage>. doi: <pub-id pub-id-type="doi">10.1093/nar/gky889</pub-id>, PMID: <pub-id pub-id-type="pmid">30295871</pub-id></citation>
</ref>
<ref id="ref41">
<label>41.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Rohart</surname> <given-names>F</given-names></name> <name><surname>Gautier</surname> <given-names>B</given-names></name> <name><surname>Singh</surname> <given-names>A</given-names></name> <name><surname>L&#x00EA; Cao</surname> <given-names>K-A</given-names></name></person-group>. <article-title>mixOMICS: an R package for 'OMICS feature selection and multiple data integration</article-title>. <source>PLoS Comput Biol</source>. (<year>2017</year>) <volume>13</volume>:<fpage>e1005752</fpage>. doi: <pub-id pub-id-type="doi">10.1371/journal.pcbi.1005752</pub-id>, PMID: <pub-id pub-id-type="pmid">29099853</pub-id></citation>
</ref>
<ref id="ref42">
<label>42.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Th&#x00E9;venot</surname> <given-names>EA</given-names></name> <name><surname>Roux</surname> <given-names>A</given-names></name> <name><surname>Xu</surname> <given-names>Y</given-names></name> <name><surname>Ezan</surname> <given-names>E</given-names></name> <name><surname>Junot</surname> <given-names>C</given-names></name></person-group>. <article-title>Analysis of the human adult urinary metabolome variations with age, body mass index, and gender by implementing a comprehensive workflow for univariate and OPLS statistical analyses</article-title>. <source>J Proteome Res</source>. (<year>2015</year>) <volume>14</volume>:<fpage>3322</fpage>&#x2013;<lpage>35</lpage>. doi: <pub-id pub-id-type="doi">10.1021/acs.jproteome.5b00354</pub-id>, PMID: <pub-id pub-id-type="pmid">26088811</pub-id></citation>
</ref>
<ref id="ref43">
<label>43.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Argelaguet</surname> <given-names>R</given-names></name> <name><surname>Velten</surname> <given-names>B</given-names></name> <name><surname>Arnol</surname> <given-names>D</given-names></name> <name><surname>Dietrich</surname> <given-names>S</given-names></name> <name><surname>Zenz</surname> <given-names>T</given-names></name> <name><surname>Marioni</surname> <given-names>JC</given-names></name> <etal/></person-group>. <article-title>Multi-omics factor analysis&#x2014;a framework for unsupervised integration of multi-omics data sets</article-title>. <source>Mol Syst Biol</source>. (<year>2018</year>) <volume>14</volume>:<fpage>e8124</fpage>. doi: <pub-id pub-id-type="doi">10.15252/msb.20178124</pub-id>, PMID: <pub-id pub-id-type="pmid">29925568</pub-id></citation>
</ref>
<ref id="ref44">
<label>44.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Picard</surname> <given-names>M</given-names></name> <name><surname>Scott-Boyer</surname> <given-names>MP</given-names></name> <name><surname>Bodein</surname> <given-names>A</given-names></name> <name><surname>P&#x00E9;rin</surname> <given-names>O</given-names></name> <name><surname>Droit</surname> <given-names>A</given-names></name></person-group>. <article-title>Integration strategies of multi-omics data for machine learning analysis</article-title>. <source>Comput Struct Biotechnol J</source>. (<year>2021</year>) <volume>19</volume>:<fpage>3735</fpage>&#x2013;<lpage>46</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.csbj.2021.06.030</pub-id>, PMID: <pub-id pub-id-type="pmid">34285775</pub-id></citation>
</ref>
<ref id="ref45">
<label>45.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Argelaguet</surname> <given-names>R</given-names></name> <name><surname>Arnol</surname> <given-names>D</given-names></name> <name><surname>Bredikhin</surname> <given-names>D</given-names></name> <name><surname>Deloro</surname> <given-names>Y</given-names></name> <name><surname>Velten</surname> <given-names>B</given-names></name> <name><surname>Marioni</surname> <given-names>JC</given-names></name> <etal/></person-group>. <article-title>MOFA+: a statistical framework for comprehensive integration of multi-modal single-cell data</article-title>. <source>Genome Biol</source>. (<year>2020</year>) <volume>21</volume>:<fpage>111</fpage>&#x2013;<lpage>7</lpage>. doi: <pub-id pub-id-type="doi">10.1186/s13059-020-02015-1</pub-id>, PMID: <pub-id pub-id-type="pmid">32393329</pub-id></citation>
</ref>
<ref id="ref46">
<label>46.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Koutrouli</surname> <given-names>M</given-names></name> <name><surname>Karatzas</surname> <given-names>E</given-names></name> <name><surname>Paez-Espino</surname> <given-names>D</given-names></name> <name><surname>Pavlopoulos</surname> <given-names>GA</given-names></name></person-group>. <article-title>A guide to conquer the biological network era using graph theory</article-title>. <source>Front Bioeng Biotechnol</source>. (<year>2020</year>) <volume>8</volume>:<fpage>34</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fbioe.2020.00034</pub-id>, PMID: <pub-id pub-id-type="pmid">32083072</pub-id></citation>
</ref>
<ref id="ref47">
<label>47.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Freitas</surname> <given-names>JANLF</given-names></name> <name><surname>Bischof</surname> <given-names>O</given-names></name></person-group>. <article-title>Dynamic modeling of the cellular senescence gene regulatory network</article-title>. <source>Heliyon</source>. (<year>2023</year>) <volume>9</volume>:<fpage>e14007</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.heliyon.2023.e14007</pub-id>, PMID: <pub-id pub-id-type="pmid">36938415</pub-id></citation>
</ref>
<ref id="ref48">
<label>48.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Froese</surname> <given-names>N</given-names></name> <name><surname>Cordero</surname> <given-names>J</given-names></name> <name><surname>Abouissa</surname> <given-names>A</given-names></name> <name><surname>Trogisch</surname> <given-names>FA</given-names></name> <name><surname>Grein</surname> <given-names>S</given-names></name> <name><surname>Szaroszyk</surname> <given-names>M</given-names></name> <etal/></person-group>. <article-title>Analysis of myocardial cellular gene expression during pressure overload reveals matrix based functional intercellular communication</article-title>. <source>iScience</source>. (<year>2022</year>) <volume>25</volume>:<fpage>103965</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.isci.2022.103965</pub-id>, PMID: <pub-id pub-id-type="pmid">35281736</pub-id></citation>
</ref>
<ref id="ref49">
<label>49.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Malod-Dognin</surname> <given-names>N</given-names></name> <name><surname>Petschnigg</surname> <given-names>J</given-names></name> <name><surname>Windels</surname> <given-names>SF</given-names></name> <name><surname>Povh</surname> <given-names>J</given-names></name> <name><surname>Hemingway</surname> <given-names>H</given-names></name> <name><surname>Ketteler</surname> <given-names>R</given-names></name> <etal/></person-group>. <article-title>Towards a data-integrated cell</article-title>. <source>Nat Commun</source>. (<year>2019</year>) <volume>10</volume>:<fpage>805</fpage>. doi: <pub-id pub-id-type="doi">10.1038/s41467-019-08797-8</pub-id>, PMID: <pub-id pub-id-type="pmid">30778056</pub-id></citation>
</ref>
<ref id="ref50">
<label>50.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Son</surname> <given-names>M</given-names></name> <name><surname>Frank</surname> <given-names>T</given-names></name> <name><surname>Holst-Hansen</surname> <given-names>T</given-names></name> <name><surname>Wang</surname> <given-names>AG</given-names></name> <name><surname>Junkin</surname> <given-names>M</given-names></name> <name><surname>Kashaf</surname> <given-names>SS</given-names></name> <etal/></person-group>. <article-title>Spatiotemporal NF-&#x03BA;B dynamics encodes the position, amplitude, and duration of local immune inputs</article-title>. <source>Sci Adv</source>. (<year>2022</year>) <volume>8</volume>:<fpage>eabn6240</fpage>. doi: <pub-id pub-id-type="doi">10.1126/sciadv.abn6240</pub-id>, PMID: <pub-id pub-id-type="pmid">36044569</pub-id></citation>
</ref>
<ref id="ref51">
<label>51.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Mart&#x00ED;nez-Zamudio</surname> <given-names>RI</given-names></name> <name><surname>Roux</surname> <given-names>PF</given-names></name> <name><surname>de Freitas</surname> <given-names>JAN</given-names></name> <name><surname>Robinson</surname> <given-names>L</given-names></name> <name><surname>Dor&#x00E9;</surname> <given-names>G</given-names></name> <name><surname>Sun</surname> <given-names>B</given-names></name> <etal/></person-group>. <article-title>AP-1 imprints a reversible transcriptional programme of senescent cells</article-title>. <source>Nat Cell Biol</source>. (<year>2020</year>) <volume>22</volume>:<fpage>842</fpage>&#x2013;<lpage>55</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41556-020-0529-5</pub-id>, PMID: <pub-id pub-id-type="pmid">32514071</pub-id></citation>
</ref>
<ref id="ref52">
<label>52.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Mart&#x00ED;nez-Zamudio</surname> <given-names>RI</given-names></name> <name><surname>Stefa</surname> <given-names>A</given-names></name> <name><surname>Freitas</surname> <given-names>JANLF</given-names></name> <name><surname>Vasilopoulos</surname> <given-names>T</given-names></name> <name><surname>Simpson</surname> <given-names>M</given-names></name> <name><surname>Dor&#x00E9;</surname> <given-names>G</given-names></name> <etal/></person-group>. <article-title>Escape from oncogene-induced senescence is controlled by POU2F2 and memorized by chromatin scars</article-title>. <source>Cell Genomics</source>. (<year>2023</year>) <volume>3</volume>:<fpage>100293</fpage>&#x2013;<lpage>30</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.xgen.2023.100293</pub-id>, PMID: <pub-id pub-id-type="pmid">37082139</pub-id></citation>
</ref>
<ref id="ref53">
<label>53.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Thigamine</surname> <given-names>K</given-names></name> <name><surname>Freitas</surname> <given-names>JANLF</given-names></name> <name><surname>Nemazanyy</surname> <given-names>I</given-names></name> <name><surname>Bankol&#x00E9;</surname> <given-names>A</given-names></name> <name><surname>Benarroch-Popivker</surname> <given-names>D</given-names></name> <name><surname>Brodesser</surname> <given-names>S</given-names></name> <etal/></person-group>. <article-title>A homeostatic switch causing glycerol-3-phosphate and phosphoethanolamine accumulation triggers senescence by rewiring lipid metabolism</article-title>. <source>Nat Metab</source>. (<year>2024</year>) <volume>6</volume>:<fpage>323</fpage>&#x2013;<lpage>42</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s42255-023-00972-y</pub-id></citation>
</ref>
<ref id="ref54">
<label>54.</label>
<citation citation-type="journal"><person-group person-group-type="author">
<collab id="coll1">The Tabula Muris Consortium</collab>
</person-group>. <article-title>A single-cell transcriptomic atlas characterizes ageing tissues in the mouse</article-title>. <source>Nature</source>. (<year>2020</year>) <volume>583</volume>:<fpage>590</fpage>&#x2013;<lpage>5</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41586-020-2496-1</pub-id>, PMID: <pub-id pub-id-type="pmid">32669714</pub-id></citation>
</ref>
<ref id="ref55">
<label>55.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Uyar</surname> <given-names>B</given-names></name> <name><surname>Palmer</surname> <given-names>D</given-names></name> <name><surname>Kowald</surname> <given-names>A</given-names></name> <name><surname>Escobar</surname> <given-names>HM</given-names></name> <name><surname>Barrantes</surname> <given-names>I</given-names></name> <name><surname>M&#x00F6;ller</surname> <given-names>S</given-names></name> <etal/></person-group>. <article-title>Single-cell analyses of aging, inflammation and senescence</article-title>. <source>Ageing Res Rev</source>. (<year>2020</year>) <volume>64</volume>:<fpage>101156</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.arr.2020.101156</pub-id>, PMID: <pub-id pub-id-type="pmid">32949770</pub-id></citation>
</ref>
<ref id="ref56">
<label>56.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Wang</surname> <given-names>R</given-names></name> <name><surname>Zhang</surname> <given-names>P</given-names></name> <name><surname>Wang</surname> <given-names>J</given-names></name> <name><surname>Ma</surname> <given-names>L</given-names></name> <name><surname>Weigao</surname> <given-names>E</given-names></name> <name><surname>Suo</surname> <given-names>S</given-names></name> <etal/></person-group>. <article-title>Construction of a cross-species cell landscape at single-cell level</article-title>. <source>Nucleic Acids Res</source>. (<year>2023</year>) <volume>51</volume>:<fpage>501</fpage>&#x2013;<lpage>16</lpage>. doi: <pub-id pub-id-type="doi">10.1093/nar/gkac633</pub-id></citation>
</ref>
<ref id="ref57">
<label>57.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Yu</surname> <given-names>Q</given-names></name> <name><surname>Jiang</surname> <given-names>M</given-names></name> <name><surname>Wu</surname> <given-names>L</given-names></name></person-group>. <article-title>Spatial transcriptomics technology in cancer research</article-title>. <source>Front Oncol</source>. (<year>2022</year>) <volume>12</volume>:<fpage>1019111</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fonc.2022.1019111</pub-id>, PMID: <pub-id pub-id-type="pmid">36313703</pub-id></citation>
</ref>
<ref id="ref58">
<label>58.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Klaus</surname> <given-names>VS</given-names></name> <name><surname>Schriever</surname> <given-names>SC</given-names></name> <name><surname>Kuhn</surname> <given-names>JMM</given-names></name> <name><surname>Peter</surname> <given-names>A</given-names></name> <name><surname>Irmler</surname> <given-names>M</given-names></name> <name><surname>Tokarz</surname> <given-names>J</given-names></name> <etal/></person-group>. <article-title>Correlation guided network integration (CoNI) reveals novel genes affecting hepatic metabolism</article-title>. <source>Mol Metab</source>. (<year>2021</year>) <volume>53</volume>:<fpage>101295</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.molmet.2021.101295</pub-id>, PMID: <pub-id pub-id-type="pmid">34271221</pub-id></citation>
</ref>
<ref id="ref59">
<label>59.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Huang</surname> <given-names>X</given-names></name> <name><surname>Bajpai</surname> <given-names>AK</given-names></name> <name><surname>Sun</surname> <given-names>J</given-names></name> <name><surname>Xu</surname> <given-names>F</given-names></name> <name><surname>Lu</surname> <given-names>L</given-names></name> <name><surname>Yousefi</surname> <given-names>S</given-names></name></person-group>. <article-title>A new gene-scoring method for uncovering novel glaucoma-related genes using non-negative matrix factorization based on RNA-seq data</article-title>. <source>Front Genet</source>. (<year>2023a</year>) <volume>14</volume>:<fpage>1204909</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fgene.2023.1204909</pub-id>, PMID: <pub-id pub-id-type="pmid">37377596</pub-id></citation>
</ref>
<ref id="ref60">
<label>60.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Zeng</surname> <given-names>Z</given-names></name> <name><surname>Vo</surname> <given-names>AH</given-names></name> <name><surname>Mao</surname> <given-names>C</given-names></name> <name><surname>Clare</surname> <given-names>SE</given-names></name> <name><surname>Khan</surname> <given-names>SA</given-names></name> <name><surname>Luo</surname> <given-names>Y</given-names></name></person-group>. <article-title>Cancer classification and pathway discovery using non-negative matrix factorization</article-title>. <source>J Biomed Inform</source>. (<year>2019</year>) <volume>96</volume>:<fpage>103247</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.jbi.2019.103247</pub-id>, PMID: <pub-id pub-id-type="pmid">31271844</pub-id></citation>
</ref>
<ref id="ref61">
<label>61.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Huang</surname> <given-names>X</given-names></name> <name><surname>Song</surname> <given-names>C</given-names></name> <name><surname>Zhang</surname> <given-names>G</given-names></name> <name><surname>Li</surname> <given-names>Y</given-names></name> <name><surname>Zhao</surname> <given-names>Y</given-names></name> <name><surname>Zhang</surname> <given-names>Q</given-names></name> <etal/></person-group>. <article-title>scGRN: a comprehensive single-cell gene regulatory network platform of human and mouse</article-title>. <source>Nucleic Acids Res</source>. (<year>2023b</year>) <volume>52</volume>:<fpage>D293</fpage>&#x2013;<lpage>303</lpage>. doi: <pub-id pub-id-type="doi">10.1093/nar/gkad885</pub-id>, PMID: <pub-id pub-id-type="pmid">37889053</pub-id></citation>
</ref>
<ref id="ref62">
<label>62.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Ren</surname> <given-names>X</given-names></name> <name><surname>Kuan</surname> <given-names>PF</given-names></name></person-group>. <article-title>RNAAgeCalc: A multi-tissue transcriptional age calculator</article-title>. <source>PLoS One</source>. (<year>2020</year>) <volume>15</volume>:<fpage>e0237006</fpage>. doi: <pub-id pub-id-type="doi">10.1371/journal.pone.0237006</pub-id>, PMID: <pub-id pub-id-type="pmid">32750074</pub-id></citation>
</ref>
<ref id="ref63">
<label>63.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Sachs</surname> <given-names>K</given-names></name> <name><surname>Perez</surname> <given-names>O</given-names></name> <name><surname>Pe'er</surname> <given-names>D</given-names></name> <name><surname>Lauffenburger</surname> <given-names>DA</given-names></name> <name><surname>Nolan</surname> <given-names>GP</given-names></name></person-group>. <article-title>Causal protein-signaling networks derived from multiparameter single-cell data</article-title>. <source>Science</source>. (<year>2005</year>) <volume>308</volume>:<fpage>523</fpage>&#x2013;<lpage>9</lpage>. doi: <pub-id pub-id-type="doi">10.1126/science.1105809</pub-id>, PMID: <pub-id pub-id-type="pmid">15845847</pub-id></citation>
</ref>
<ref id="ref64">
<label>64.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Lachmann</surname> <given-names>A</given-names></name> <name><surname>Giorgi</surname> <given-names>FM</given-names></name> <name><surname>Lopez</surname> <given-names>G</given-names></name> <name><surname>Califano</surname> <given-names>A</given-names></name></person-group>. <article-title>ARACNe-AP: gene network reverse engineering through adaptive partitioning inference of mutual information</article-title>. <source>Bioinformatics</source>. (<year>2016</year>) <volume>32</volume>:<fpage>2233</fpage>&#x2013;<lpage>5</lpage>. doi: <pub-id pub-id-type="doi">10.1093/bioinformatics/btw216</pub-id>, PMID: <pub-id pub-id-type="pmid">27153652</pub-id></citation>
</ref>
<ref id="ref65">
<label>65.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Rao</surname> <given-names>M</given-names></name> <name><surname>Wang</surname> <given-names>X</given-names></name> <name><surname>Guo</surname> <given-names>G</given-names></name> <name><surname>Wang</surname> <given-names>L</given-names></name> <name><surname>Chen</surname> <given-names>S</given-names></name> <name><surname>Yin</surname> <given-names>P</given-names></name> <etal/></person-group>. <article-title>Resolving the intertwining of inflammation and fibrosis in human heart failure at single-cell level</article-title>. <source>Basic Res Cardiol</source>. (<year>2021</year>) <volume>116</volume>:<fpage>55</fpage>&#x2013;<lpage>19</lpage>. doi: <pub-id pub-id-type="doi">10.1007/s00395-021-00897-1</pub-id>, PMID: <pub-id pub-id-type="pmid">34601654</pub-id></citation>
</ref>
<ref id="ref66">
<label>66.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Wang</surname> <given-names>W</given-names></name> <name><surname>Vilella</surname> <given-names>F</given-names></name> <name><surname>Alama</surname> <given-names>P</given-names></name> <name><surname>Moreno</surname> <given-names>I</given-names></name> <name><surname>Mignardi</surname> <given-names>M</given-names></name> <name><surname>Isakova</surname> <given-names>A</given-names></name> <etal/></person-group>. <article-title>Single-cell transcriptomic atlas of the human endometrium during the menstrual cycle</article-title>. <source>Nat Med</source>. (<year>2020</year>) <volume>26</volume>:<fpage>1644</fpage>&#x2013;<lpage>53</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41591-020-1040-z</pub-id>, PMID: <pub-id pub-id-type="pmid">32929266</pub-id></citation>
</ref>
<ref id="ref67">
<label>67.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Wang</surname> <given-names>K</given-names></name> <name><surname>Saito</surname> <given-names>M</given-names></name> <name><surname>Bisikirska</surname> <given-names>BC</given-names></name> <name><surname>Alvarez</surname> <given-names>MJ</given-names></name> <name><surname>Lim</surname> <given-names>WK</given-names></name> <name><surname>Rajbhandari</surname> <given-names>P</given-names></name> <etal/></person-group>. <article-title>Genome-wide identification of post-translational modulators of transcription factor activity in human B cells</article-title>. <source>Nat Biotechnol</source>. (<year>2009</year>) <volume>27</volume>:<fpage>829</fpage>&#x2013;<lpage>37</lpage>. doi: <pub-id pub-id-type="doi">10.1038/nbt.1563</pub-id>, PMID: <pub-id pub-id-type="pmid">19741643</pub-id></citation>
</ref>
<ref id="ref68">
<label>68.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Okada</surname> <given-names>D</given-names></name> <name><surname>Cheng</surname> <given-names>JH</given-names></name> <name><surname>Zheng</surname> <given-names>C</given-names></name> <name><surname>Kumaki</surname> <given-names>T</given-names></name> <name><surname>Yamada</surname> <given-names>R</given-names></name></person-group>. <article-title>Data-driven identification and classification of nonlinear aging patterns reveals the landscape of associations between DNA methylation and aging</article-title>. <source>Hum Genomics</source>. (<year>2023</year>) <volume>17</volume>:<fpage>8</fpage>. doi: <pub-id pub-id-type="doi">10.1186/s40246-023-00453-z</pub-id></citation>
</ref>
<ref id="ref69">
<label>69.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Konrath</surname> <given-names>F</given-names></name> <name><surname>Willenbrock</surname> <given-names>M</given-names></name> <name><surname>Busse</surname> <given-names>D</given-names></name> <name><surname>Scheidereit</surname> <given-names>C</given-names></name> <name><surname>Wolf</surname> <given-names>J</given-names></name></person-group>. <article-title>A computational model of the DNA damage-induced IKK/NF-&#x03BA;B pathway reveals a critical dependence on irradiation dose and PARP-1</article-title>. <source>BioRxiv</source>. (<year>2023</year>) <volume>26</volume>:<fpage>107917</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.isci.2023.107917</pub-id>, PMID: <pub-id pub-id-type="pmid">37817938</pub-id></citation>
</ref>
<ref id="ref70">
<label>70.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Li</surname> <given-names>W</given-names></name> <name><surname>Zhao</surname> <given-names>L</given-names></name> <name><surname>Wang</surname> <given-names>J</given-names></name></person-group>. <article-title>Searching for the mechanisms of mammalian cellular aging through underlying gene regulatory networks</article-title>. <source>Front Genet</source>. (<year>2020</year>) <volume>11</volume>:<fpage>593</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fgene.2020.00593</pub-id>, PMID: <pub-id pub-id-type="pmid">32714367</pub-id></citation>
</ref>
<ref id="ref71">
<label>71.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Kurata</surname> <given-names>H</given-names></name> <name><surname>Maeda</surname> <given-names>K</given-names></name> <name><surname>Onaka</surname> <given-names>T</given-names></name> <name><surname>Takata</surname> <given-names>T</given-names></name></person-group>. <article-title>BioFNet: biological functional network database for analysis and synthesis of biological systems</article-title>. <source>Brief Bioinform</source>. (<year>2014</year>) <volume>15</volume>:<fpage>699</fpage>&#x2013;<lpage>709</lpage>. doi: <pub-id pub-id-type="doi">10.1093/bib/bbt048</pub-id>, PMID: <pub-id pub-id-type="pmid">23894104</pub-id></citation>
</ref>
<ref id="ref72">
<label>72.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Dyson</surname> <given-names>J</given-names></name> <name><surname>Villella-Bressan</surname> <given-names>R</given-names></name> <name><surname>Webb</surname> <given-names>G</given-names></name></person-group>. <article-title>A spatial model of tumor growth with cell age, cell size, and mutation of cell phenotypes</article-title>. <source>Math Modell Nat Phenom</source>. (<year>2007</year>) <volume>2</volume>:<fpage>69</fpage>&#x2013;<lpage>100</lpage>. doi: <pub-id pub-id-type="doi">10.1051/mmnp:2007004</pub-id></citation>
</ref>
<ref id="ref73">
<label>73.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Kuppe</surname> <given-names>C</given-names></name> <name><surname>Ramirez Flores</surname> <given-names>RO</given-names></name> <name><surname>Li</surname> <given-names>Z</given-names></name> <name><surname>Hayat</surname> <given-names>S</given-names></name> <name><surname>Levinson</surname> <given-names>RT</given-names></name> <name><surname>Liao</surname> <given-names>X</given-names></name> <etal/></person-group>. <article-title>Spatial multi-omic map of human myocardial infarction</article-title>. <source>Nature</source>. (<year>2022</year>) <volume>608</volume>:<fpage>766</fpage>&#x2013;<lpage>77</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41586-022-05060-x</pub-id>, PMID: <pub-id pub-id-type="pmid">35948637</pub-id></citation>
</ref>
<ref id="ref74">
<label>74.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Driever</surname> <given-names>W</given-names></name> <name><surname>N&#x00FC;sslein-Volhard</surname> <given-names>C</given-names></name></person-group>. <article-title>The bicoid protein determines the position in the Drosophila embryo in a concentration-dependent manner</article-title>. <source>Cell</source>. (<year>1988</year>) <volume>54</volume>:<fpage>95</fpage>&#x2013;<lpage>104</lpage>. doi: <pub-id pub-id-type="doi">10.1016/0092-8674(88)90183-3</pub-id></citation>
</ref>
<ref id="ref75">
<label>75.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Mousavi</surname> <given-names>R</given-names></name> <name><surname>Lobo</surname> <given-names>D</given-names></name></person-group>. <article-title>Automatic design of gene regulatory mechanisms for spatial pattern formation</article-title>. <source>NPJ Syst Biol Appl</source>. (<year>2024</year>) <volume>10</volume>:<fpage>35</fpage>. doi: <pub-id pub-id-type="doi">10.1038/s41540-024-00361-5</pub-id>, PMID: <pub-id pub-id-type="pmid">38565850</pub-id></citation>
</ref>
<ref id="ref76">
<label>76.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Grodstein</surname> <given-names>J</given-names></name> <name><surname>McMillen</surname> <given-names>P</given-names></name> <name><surname>Levin</surname> <given-names>M</given-names></name></person-group>. <article-title>Closing the loop on morphogenesis: a mathematical model of morphogenesis by closed-loop reaction-diffusion</article-title>. <source>Front Cell Dev Biol</source>. (<year>2023</year>) <volume>11</volume>:<fpage>1087650</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fcell.2023.1087650</pub-id>, PMID: <pub-id pub-id-type="pmid">37645245</pub-id></citation>
</ref>
<ref id="ref77">
<label>77.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Turing</surname> <given-names>AM</given-names></name>
</person-group>. <article-title>The chemical basis of morphogenesis</article-title>. <source>Phil Trans R Soc Lond B</source>. (<year>1952</year>) <volume>237</volume>:<fpage>37</fpage>&#x2013;<lpage>72</lpage>. doi: <pub-id pub-id-type="doi">10.1098/rstb.1952.0012</pub-id></citation>
</ref>
<ref id="ref78">
<label>78.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Reinitz</surname> <given-names>J</given-names></name>
</person-group>. <article-title>Pattern formation</article-title>. <source>Nature</source>. (<year>2012</year>) <volume>482</volume>:<fpage>464</fpage>&#x2013;<lpage>4</lpage>. doi: <pub-id pub-id-type="doi">10.1038/482464a</pub-id></citation>
</ref>
<ref id="ref79">
<label>79.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Maini</surname> <given-names>PK</given-names></name> <name><surname>Woolley</surname> <given-names>TE</given-names></name> <name><surname>Baker</surname> <given-names>RE</given-names></name> <name><surname>Gaffney</surname> <given-names>EA</given-names></name> <name><surname>Lee</surname> <given-names>SS</given-names></name></person-group>. <article-title>Turing's model for biological pattern formation and the robustness problem</article-title>. <source>Interface Focus</source>. (<year>2012</year>) <volume>2</volume>:<fpage>487</fpage>&#x2013;<lpage>96</lpage>. doi: <pub-id pub-id-type="doi">10.1098/rsfs.2011.0113</pub-id>, PMID: <pub-id pub-id-type="pmid">23919129</pub-id></citation>
</ref>
<ref id="ref80">
<label>80.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Lagemann</surname> <given-names>K</given-names></name> <name><surname>Lagemann</surname> <given-names>C</given-names></name> <name><surname>Taschler</surname> <given-names>B</given-names></name> <name><surname>Mukherjee</surname> <given-names>S</given-names></name></person-group>. <article-title>Deep learning of causal structures in high dimensions under data limitations</article-title>. <source>Nat Mach Intell</source>. (<year>2023</year>) <volume>5</volume>:<fpage>1306</fpage>&#x2013;<lpage>16</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s42256-023-00744-z</pub-id></citation>
</ref>
<ref id="ref81">
<label>81.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Russakovsky</surname> <given-names>O</given-names></name> <name><surname>Deng</surname> <given-names>J</given-names></name> <name><surname>Su</surname> <given-names>H</given-names></name> <name><surname>Krause</surname> <given-names>J</given-names></name> <name><surname>Satheesh</surname> <given-names>S</given-names></name> <name><surname>Ma</surname> <given-names>S</given-names></name> <etal/></person-group>. <article-title>Imagenet large scale visual recognition challenge</article-title>. <source>Int J Comput Vis</source>. (<year>2015</year>) <volume>115</volume>:<fpage>211</fpage>&#x2013;<lpage>52</lpage>. doi: <pub-id pub-id-type="doi">10.1007/s11263-015-0816-y</pub-id></citation>
</ref>
<ref id="ref82">
<label>82.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Chen</surname> <given-names>Y</given-names></name> <name><surname>Wang</surname> <given-names>Y</given-names></name> <name><surname>Chen</surname> <given-names>Y</given-names></name> <name><surname>Cheng</surname> <given-names>Y</given-names></name> <name><surname>Wei</surname> <given-names>Y</given-names></name> <name><surname>Li</surname> <given-names>Y</given-names></name> <etal/></person-group>, <article-title>Deep autoencoder for interpretable tissue-adaptive deconvolution and cell-type-specific gene analysis</article-title>. <source>Nat. Commun.</source> (<year>2022</year>) <volume>13</volume>, <fpage>6735</fpage>. doi: <pub-id pub-id-type="doi">10.1038/s41467-022-34550-9</pub-id></citation>
</ref>
<ref id="ref83">
<label>83.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Qiu</surname> <given-names>W</given-names></name> <name><surname>Chen</surname> <given-names>H</given-names></name> <name><surname>Kaeberlein</surname> <given-names>M</given-names></name> <name><surname>Lee</surname> <given-names>S. I.</given-names></name></person-group> <article-title>ExplaiNAble BioLogical Age (ENABL Age): an artificial intelligence framework for interpretable biological age</article-title>. <source>The Lancet Healthy Longevity.</source> (<year>2023</year>) <volume>4</volume>, <fpage>e711</fpage>&#x2013;<lpage>e723</lpage>. doi: <pub-id pub-id-type="doi">10.1016/S2666-7568(23)00189-7</pub-id></citation>
</ref>
<ref id="ref84">
<label>84.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Holzscheck</surname> <given-names>N</given-names></name> <name><surname>Falckenhayn</surname> <given-names>C</given-names></name> <name><surname>S&#x00F6;hle</surname> <given-names>J</given-names></name> <name><surname>Kristof</surname> <given-names>B</given-names></name> <name><surname>Siegner</surname> <given-names>R</given-names></name> <name><surname>Werner</surname> <given-names>A</given-names></name> <etal/></person-group>. <article-title>Modeling transcriptomic age using knowledge-rimed artificial neural networks</article-title>. <source>npj Aging and Mechanisms of Disease</source>. (<year>2021</year>) <volume>7</volume>, <fpage>15</fpage>. doi: <pub-id pub-id-type="doi">10.1038/s41514-021-00068-5</pub-id></citation>
</ref>
<ref id="ref85">
<label>85.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Tao</surname> <given-names>W</given-names></name> <name><surname>Yu</surname> <given-names>Z</given-names></name> <name><surname>Han</surname> <given-names>JDJ</given-names></name></person-group>. <article-title>Single-cell senescence identification reveals senescence heterogeneity, trajectory, and modulators</article-title>. <source>Cell Metab</source>. (<year>2024</year>) <volume>36</volume>:<fpage>1126</fpage>&#x2013;<lpage>1143</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.cmet.2024.03.009</pub-id>, PMID: <pub-id pub-id-type="pmid">38604170</pub-id></citation>
</ref>
<ref id="ref86">
<label>86.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Vazquez</surname> <given-names>J</given-names></name> <name><surname>Facelli</surname> <given-names>J. C.</given-names></name></person-group>, <article-title>Conformal prediction in clinical medical sciences</article-title>. <source>J. Healthc. Inform. Res.</source> (<year>2022</year>) <volume>6</volume>, <fpage>241</fpage>&#x2013;<lpage>252</lpage>. doi: <pub-id pub-id-type="doi">10.1007/s41666-021-00113-8</pub-id></citation>
</ref>
<ref id="ref87">
<label>87.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Sanchez-Vega</surname> <given-names>F</given-names></name> <name><surname>Mina</surname> <given-names>M</given-names></name> <name><surname>Armenia</surname> <given-names>J</given-names></name> <name><surname>Chatila</surname> <given-names>WK</given-names></name> <name><surname>Luna</surname> <given-names>A</given-names></name> <name><surname>La</surname> <given-names>KC</given-names></name> <etal/></person-group>. <article-title>Oncogenic signaling pathways in the cancer genome atlas</article-title>. <source>Cell</source>. (<year>2018</year>) <volume>173</volume>:<fpage>321</fpage>&#x2013;<lpage>337.e10</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.cell.2018.03.035</pub-id>, PMID: <pub-id pub-id-type="pmid">29625050</pub-id></citation>
</ref>
<ref id="ref88">
<label>88.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Ding</surname> <given-names>L</given-names></name> <name><surname>Bailey</surname> <given-names>MH</given-names></name> <name><surname>Porta-Pardo</surname> <given-names>E</given-names></name> <name><surname>Thorsson</surname> <given-names>V</given-names></name> <name><surname>Colaprico</surname> <given-names>A</given-names></name> <name><surname>Bertrand</surname> <given-names>D</given-names></name> <etal/></person-group>. <article-title>Perspective on oncogenic processes at the end of the beginning of cancer genomics</article-title>. <source>Cell</source>. (<year>2018</year>) <volume>173</volume>:<fpage>305</fpage>&#x2013;<lpage>320.e10</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.cell.2018.03.033</pub-id>, PMID: <pub-id pub-id-type="pmid">29625049</pub-id></citation>
</ref>
<ref id="ref89">
<label>89.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Hoadley</surname> <given-names>KA</given-names></name> <name><surname>Yau</surname> <given-names>C</given-names></name> <name><surname>Hinoue</surname> <given-names>T</given-names></name> <name><surname>Wolf</surname> <given-names>DM</given-names></name> <name><surname>Lazar</surname> <given-names>AJ</given-names></name> <name><surname>Drill</surname> <given-names>E</given-names></name> <etal/></person-group>. <article-title>Cell-of-origin patterns dominate the molecular classification of 10,000 tumors from 33 types of cancer</article-title>. <source>Cell</source>. (<year>2018</year>) <volume>173</volume>:<fpage>291</fpage>&#x2013;<lpage>304.e6</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.cell.2018.03.022</pub-id>, PMID: <pub-id pub-id-type="pmid">29625048</pub-id></citation>
</ref>
<ref id="ref90">
<label>90.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Costanzo</surname> <given-names>MC</given-names></name> <name><surname>von Grotthuss</surname> <given-names>M</given-names></name> <name><surname>Massung</surname> <given-names>J</given-names></name> <name><surname>Jang</surname> <given-names>D</given-names></name> <name><surname>Caulkins</surname> <given-names>L</given-names></name> <name><surname>Koesterer</surname> <given-names>R</given-names></name> <etal/></person-group>. <article-title>The type 2 diabetes knowledge portal: an open access genetic resource dedicated to type 2 diabetes and related traits</article-title>. <source>Cell Metab</source>. (<year>2023a</year>) <volume>35</volume>:<fpage>695</fpage>&#x2013;<lpage>710.e6</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.cmet.2023.03.001</pub-id>, PMID: <pub-id pub-id-type="pmid">36963395</pub-id></citation>
</ref>
<ref id="ref91">
<label>91.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Costanzo</surname> <given-names>MC</given-names></name> <name><surname>Roselli</surname> <given-names>C</given-names></name> <name><surname>Brandes</surname> <given-names>M</given-names></name> <name><surname>Duby</surname> <given-names>M</given-names></name> <name><surname>Hoang</surname> <given-names>Q</given-names></name> <name><surname>Jang</surname> <given-names>D</given-names></name> <etal/></person-group>. <article-title>Cardiovascular disease knowledge portal: a community resource for cardiovascular disease research</article-title>. <source>Circ Genomic Precis Med</source>. (<year>2023b</year>) <volume>16</volume>:<fpage>e004181</fpage>. doi: <pub-id pub-id-type="doi">10.1161/CIRCGEN.123.004181</pub-id></citation>
</ref>
<ref id="ref92">
<label>92.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Rafikova</surname> <given-names>E</given-names></name> <name><surname>Nemirovich-Danchenko</surname> <given-names>N</given-names></name> <name><surname>Ogmen</surname> <given-names>A</given-names></name> <name><surname>Parfenenkova</surname> <given-names>A</given-names></name> <name><surname>Velikanova</surname> <given-names>A</given-names></name> <name><surname>Tikhonov</surname> <given-names>S</given-names></name> <etal/></person-group>. <article-title>Open genes&#x2014;a new comprehensive database of human genes associated with aging and longevity</article-title>. <source>Nucleic Acids Res</source>. (<year>2024</year>) <volume>52</volume>:<fpage>D950</fpage>&#x2013;<lpage>62</lpage>. doi: <pub-id pub-id-type="doi">10.1093/nar/gkad712</pub-id>, PMID: <pub-id pub-id-type="pmid">37665017</pub-id></citation>
</ref>
<ref id="ref93">
<label>93.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Zappia</surname> <given-names>L</given-names></name> <name><surname>Phipson</surname> <given-names>B</given-names></name> <name><surname>Oshlack</surname> <given-names>A</given-names></name></person-group>. <article-title>Exploring the single-cell RNA-seq analysis landscape with the scRNA-tools database</article-title>. <source>PLoS Comput Biol</source>. (<year>2018</year>) <volume>14</volume>:<fpage>e1006245</fpage>. doi: <pub-id pub-id-type="doi">10.1371/journal.pcbi.1006245</pub-id>, PMID: <pub-id pub-id-type="pmid">29939984</pub-id></citation>
</ref>
<ref id="ref94">
<label>94.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Rajewsky</surname> <given-names>N</given-names></name> <name><surname>Almouzni</surname> <given-names>G</given-names></name> <name><surname>Gorski</surname> <given-names>SA</given-names></name> <name><surname>Aerts</surname> <given-names>S</given-names></name> <name><surname>Amit</surname> <given-names>I</given-names></name> <name><surname>Bertero</surname> <given-names>MG</given-names></name> <etal/></person-group>. <article-title>LifeTime and improving European healthcare through cell-based interceptive medicine</article-title>. <source>Nature</source>. (<year>2020</year>) <volume>587</volume>:<fpage>377</fpage>&#x2013;<lpage>86</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41586-020-2715-9</pub-id>, PMID: <pub-id pub-id-type="pmid">32894860</pub-id></citation>
</ref>
<ref id="ref95">
<label>95.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Regev</surname> <given-names>A</given-names></name> <name><surname>Teichmann</surname> <given-names>SA</given-names></name> <name><surname>Lander</surname> <given-names>ES</given-names></name> <name><surname>Amit</surname> <given-names>I</given-names></name> <name><surname>Benoist</surname> <given-names>C</given-names></name> <name><surname>Birney</surname> <given-names>E</given-names></name> <etal/></person-group>. <article-title>The human cell atlas</article-title>. <source>eLife</source>. (<year>2017</year>) <volume>6</volume>:<fpage>e27041</fpage>. doi: <pub-id pub-id-type="doi">10.7554/eLife.27041</pub-id></citation>
</ref>
<ref id="ref96">
<label>96.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Thul</surname> <given-names>PJ</given-names></name> <name><surname>Lindskog</surname> <given-names>C</given-names></name></person-group>. <article-title>The human protein atlas: a spatial map of the human proteome</article-title>. <source>Protein Sci</source>. (<year>2017</year>) <volume>27</volume>:<fpage>233</fpage>&#x2013;<lpage>44</lpage>. doi: <pub-id pub-id-type="doi">10.1002/pro.3307</pub-id>, PMID: <pub-id pub-id-type="pmid">28940711</pub-id></citation>
</ref>
<ref id="ref97">
<label>97.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Karlsson</surname> <given-names>M</given-names></name> <name><surname>Zhang</surname> <given-names>C</given-names></name> <name><surname>M&#x00E9;ar</surname> <given-names>L</given-names></name> <name><surname>Zhong</surname> <given-names>W</given-names></name> <name><surname>Digre</surname> <given-names>A</given-names></name> <name><surname>Katona</surname> <given-names>B</given-names></name> <etal/></person-group>. <article-title>A single&#x2013;cell type transcriptomics map of human tissues</article-title>. <source>Sci Adv</source>. (<year>2021</year>) <volume>7</volume>:<fpage>eabh2169</fpage>. doi: <pub-id pub-id-type="doi">10.1126/sciadv.abh2169</pub-id>, PMID: <pub-id pub-id-type="pmid">34321199</pub-id></citation>
</ref>
<ref id="ref98">
<label>98.</label>
<citation citation-type="other"><person-group person-group-type="author"><collab id="coll2">Roadmap EpigenOMICS Consortium</collab><name><surname>Kundaje</surname> <given-names>A</given-names></name> <name><surname>Meuleman</surname> <given-names>W</given-names></name> <name><surname>Ernst</surname> <given-names>J</given-names></name> <name><surname>Bilenky</surname> <given-names>M</given-names></name> <name><surname>Yen</surname> <given-names>A</given-names></name> <etal/></person-group>. <article-title>Integrative analysis of 111 reference human epigenomes</article-title>. <source>Nature</source>. (<year>2015</year>) <volume>518</volume>:<fpage>317</fpage>&#x2013;<lpage>30</lpage>. doi: <pub-id pub-id-type="doi">10.1038/nature14248</pub-id>, PMID: <pub-id pub-id-type="pmid">25693563</pub-id></citation>
</ref>
<ref id="ref99">
<label>99.</label>
<citation citation-type="journal"><person-group person-group-type="author"><collab id="coll3">The Tabula Sapiens Consortium&#x002A;</collab><name><surname>Jones</surname> <given-names>RC</given-names></name> <name><surname>Karkanias</surname> <given-names>J</given-names></name> <name><surname>Krasnow</surname> <given-names>MA</given-names></name> <name><surname>Pisco</surname> <given-names>AO</given-names></name> <name><surname>Quake</surname> <given-names>SR</given-names></name> <etal/></person-group>. <article-title>The tabula sapiens: A multiple-organ, single-cell transcriptomic atlas of humans</article-title>. <source>Science</source>. (<year>2022</year>) <volume>376</volume>:<fpage>eabl4896</fpage>. doi: <pub-id pub-id-type="doi">10.1126/science.abl4896</pub-id>, PMID: <pub-id pub-id-type="pmid">35549404</pub-id></citation>
</ref>
<ref id="ref100">
<label>100.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Cusanovich</surname> <given-names>DA</given-names></name> <name><surname>Hill</surname> <given-names>AJ</given-names></name> <name><surname>Aghamirzaie</surname> <given-names>D</given-names></name> <name><surname>Daza</surname> <given-names>RM</given-names></name> <name><surname>Pliner</surname> <given-names>HA</given-names></name> <name><surname>Berletch</surname> <given-names>JB</given-names></name> <etal/></person-group>. <article-title>A single-cell atlas of in vivo mammalian chromatin accessibility</article-title>. <source>Cell</source>. (<year>2018</year>) <volume>174</volume>:<fpage>1309</fpage>&#x2013;<lpage>1324.e18</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.cell.2018.06.052</pub-id>, PMID: <pub-id pub-id-type="pmid">30078704</pub-id></citation>
</ref>
<ref id="ref101">
<label>101.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Fei</surname> <given-names>L</given-names></name> <name><surname>Chen</surname> <given-names>H</given-names></name> <name><surname>Ma</surname> <given-names>L</given-names></name> <name><surname>E</surname> <given-names>W</given-names></name> <name><surname>Wang</surname> <given-names>R</given-names></name> <name><surname>Fang</surname> <given-names>X</given-names></name> <etal/></person-group>. <article-title>Systematic identification of cell-fate regulatory programs using a single-cell atlas of mouse development</article-title>. <source>Nat Genet</source>. (<year>2022</year>) <volume>54</volume>:<fpage>1051</fpage>&#x2013;<lpage>61</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41588-022-01118-8</pub-id>, PMID: <pub-id pub-id-type="pmid">35817981</pub-id></citation>
</ref>
<ref id="ref102">
<label>102.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Jiang</surname> <given-names>M</given-names></name> <name><surname>Xiao</surname> <given-names>Y</given-names></name> <name><surname>Ma</surname> <given-names>L</given-names></name> <name><surname>Wang</surname> <given-names>J</given-names></name> <name><surname>Chen</surname> <given-names>H</given-names></name> <name><surname>Gao</surname> <given-names>C</given-names></name> <etal/></person-group>. <article-title>Characterization of the zebrafish cell landscape at single-cell resolution</article-title>. <source>Front Cell Dev Biol</source>. (<year>2021</year>) <volume>9</volume>:<fpage>743421</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fcell.2021.743421</pub-id>, PMID: <pub-id pub-id-type="pmid">34660600</pub-id></citation>
</ref>
<ref id="ref103">
<label>103.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Wagner</surname> <given-names>DE</given-names></name> <name><surname>Weinreb</surname> <given-names>C</given-names></name> <name><surname>Collins</surname> <given-names>ZM</given-names></name> <name><surname>Briggs</surname> <given-names>JA</given-names></name> <name><surname>Megason</surname> <given-names>SG</given-names></name> <name><surname>Klein</surname> <given-names>AM</given-names></name></person-group>. <article-title>Single-cell mapping of gene expression landscapes and lineage in the zebrafish embryo</article-title>. <source>Science</source>. (<year>2018</year>) <volume>360</volume>:<fpage>981</fpage>&#x2013;<lpage>7</lpage>. doi: <pub-id pub-id-type="doi">10.1126/science.aar4362</pub-id>, PMID: <pub-id pub-id-type="pmid">29700229</pub-id></citation>
</ref>
<ref id="ref104">
<label>104.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Roux</surname> <given-names>AE</given-names></name> <name><surname>Yuan</surname> <given-names>H</given-names></name> <name><surname>Podshivalova</surname> <given-names>K</given-names></name> <name><surname>Hendrickson</surname> <given-names>D</given-names></name> <name><surname>Kerr</surname> <given-names>R</given-names></name> <name><surname>Kenyon</surname> <given-names>C</given-names></name> <etal/></person-group>. <article-title>Individual cell types in <italic>C. elegans</italic> age differently and activate distinct cell-protective responses</article-title>. <source>Cell Rep</source>. (<year>2023</year>) <volume>42</volume>:<fpage>112902</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.celrep.2023.112902</pub-id>, PMID: <pub-id pub-id-type="pmid">37531250</pub-id></citation>
</ref>
<ref id="ref105">
<label>105.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Packer</surname> <given-names>JS</given-names></name> <name><surname>Zhu</surname> <given-names>Q</given-names></name> <name><surname>Huynh</surname> <given-names>C</given-names></name> <name><surname>Sivaramakrishnan</surname> <given-names>P</given-names></name> <name><surname>Preston</surname> <given-names>E</given-names></name> <name><surname>Dueck</surname> <given-names>H</given-names></name> <etal/></person-group>. <article-title>A lineage-resolved molecular atlas of <italic>C. elegans</italic> embryogenesis at single-cell resolution</article-title>. <source>Science</source>. (<year>2019</year>) <volume>365</volume>:<fpage>eaax1971</fpage>. doi: <pub-id pub-id-type="doi">10.1126/science.aax1971</pub-id>, PMID: <pub-id pub-id-type="pmid">31488706</pub-id></citation>
</ref>
<ref id="ref106">
<label>106.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Siebert</surname> <given-names>S</given-names></name> <name><surname>Farrell</surname> <given-names>JA</given-names></name> <name><surname>Cazet</surname> <given-names>JF</given-names></name> <name><surname>Abeykoon</surname> <given-names>Y</given-names></name> <name><surname>Primack</surname> <given-names>AS</given-names></name> <name><surname>Schnitzler</surname> <given-names>CE</given-names></name> <etal/></person-group>. <article-title>Stem cell differentiation trajectories in Hydra resolved at single-cell resolution</article-title>. <source>Science</source>. (<year>2019</year>) <volume>365</volume>:<fpage>eaav9314</fpage>. doi: <pub-id pub-id-type="doi">10.1126/science.aav9314</pub-id>, PMID: <pub-id pub-id-type="pmid">31346039</pub-id></citation>
</ref>
<ref id="ref107">
<label>107.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Plass</surname> <given-names>M</given-names></name> <name><surname>Solana</surname> <given-names>J</given-names></name> <name><surname>Wolf</surname> <given-names>FA</given-names></name> <name><surname>Ayoub</surname> <given-names>S</given-names></name> <name><surname>Misios</surname> <given-names>A</given-names></name> <name><surname>Gla&#x017E;ar</surname> <given-names>P</given-names></name> <etal/></person-group>. <article-title>Cell type atlas and lineage tree of a whole complex animal by single-cell transcriptomics</article-title>. <source>Science</source>. (<year>2018</year>) <volume>360</volume>:<fpage>eaaq1723</fpage>. doi: <pub-id pub-id-type="doi">10.1126/science.aaq1723</pub-id>, PMID: <pub-id pub-id-type="pmid">29674432</pub-id></citation>
</ref>
<ref id="ref108">
<label>108.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Brown</surname> <given-names>T</given-names></name> <name><surname>Mann</surname> <given-names>B</given-names></name> <name><surname>Ryder</surname> <given-names>N</given-names></name> <name><surname>Subbiah</surname> <given-names>M</given-names></name> <name><surname>Kaplan</surname> <given-names>JD</given-names></name> <name><surname>Dhariwal</surname> <given-names>P</given-names></name> <etal/></person-group>,. <article-title>Language models are few-shot learners</article-title>. <source>Adv. Neural Inf. Process. Syst.</source> (<year>2020</year>) <volume>33</volume>, <fpage>1877</fpage>&#x2013;<lpage>1901</lpage>.</citation>
</ref>
<ref id="ref109">
<label>109.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Tejada-Martinez</surname> <given-names>D</given-names></name> <name><surname>De Magalh&#x00E3;es</surname> <given-names>JP</given-names></name> <name><surname>Opazo</surname> <given-names>JC</given-names></name></person-group>. <article-title>Positive selection and gene duplications in tumour suppressor genes reveal clues about how cetaceans resist cancer</article-title>. <source>Proc R Soc B</source>. (<year>2021</year>) <volume>288</volume>:<fpage>20202592</fpage>. doi: <pub-id pub-id-type="doi">10.1098/rspb.2020.2592</pub-id>, PMID: <pub-id pub-id-type="pmid">33622125</pub-id></citation>
</ref>
<ref id="ref110">
<label>110.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Mah</surname> <given-names>JL</given-names></name> <name><surname>Dunn</surname> <given-names>CW</given-names></name></person-group>. <article-title>Cell type evolution reconstruction across species through cell phylogenies of single-cell RNA sequencing data</article-title>. <source>Nature Ecology &#x0026; Evolution</source>. (<year>2024</year>) <fpage>1</fpage>&#x2013;<lpage>14</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41559-023-02281-9</pub-id></citation>
</ref>
<ref id="ref111">
<label>111.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Wang</surname> <given-names>T</given-names></name> <name><surname>Du</surname> <given-names>Y</given-names></name> <name><surname>Gong</surname> <given-names>Y</given-names></name> <name><surname>Choo</surname> <given-names>KKR</given-names></name> <name><surname>Guo</surname> <given-names>Y</given-names></name></person-group>. <article-title>Applications of federated learning in mobile health: scoping review</article-title>. <source>J Med Internet Res</source>. (<year>2023</year>) <volume>25</volume>:<fpage>e43006</fpage>. doi: <pub-id pub-id-type="doi">10.2196/43006</pub-id>, PMID: <pub-id pub-id-type="pmid">37126398</pub-id></citation>
</ref>
<ref id="ref112">
<label>112.</label>
<citation citation-type="other"><person-group person-group-type="author"><name><surname>Brunton</surname> <given-names>S. L.</given-names></name> <name><surname>Kutz</surname> <given-names>J. N.</given-names></name></person-group> (<year>2022</year>). &#x201C;<article-title>The diversity of neural networks</article-title>&#x201D;, in <source>Data-driven science and engineering: machine learning, dynamical systems, and control</source> (<publisher-loc>Cambridge</publisher-loc>: <publisher-name>Cambridge University Press</publisher-name>), <fpage>291</fpage>&#x2013;<lpage>299</lpage>. Available at: <ext-link xlink:href="https://web.archive.org/web/20230621143644/https://faculty.washington.edu/sbrunton/DataBookV2.pdf" ext-link-type="uri">https://web.archive.org/web/20230621143644/https://faculty.washington.edu/sbrunton/DataBookV2.pdf</ext-link> (Accessed April 18, 2024).</citation>
</ref>
<ref id="ref113">
<label>113.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Wassan</surname> <given-names>JT</given-names></name> <name><surname>Zheng</surname> <given-names>H</given-names></name> <name><surname>Wang</surname> <given-names>H</given-names></name></person-group>. <article-title>Role of deep learning in predicting aging-related diseases: a scoping review</article-title>. <source>Cells</source>. (<year>2021</year>) <volume>10</volume>:<fpage>2924</fpage>. doi: <pub-id pub-id-type="doi">10.3390/cells10112924</pub-id>, PMID: <pub-id pub-id-type="pmid">34831148</pub-id></citation>
</ref>
<ref id="ref114">
<label>114.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Zhang</surname> <given-names>MJ</given-names></name> <name><surname>Pisco</surname> <given-names>AO</given-names></name> <name><surname>Darmanis</surname> <given-names>S</given-names></name> <name><surname>Zou</surname> <given-names>J</given-names></name></person-group>. <article-title>Mouse aging cell atlas analysis reveals global and cell type-specific aging signatures</article-title>. <source>eLife</source>. (<year>2021</year>) <volume>10</volume>:<fpage>e62293</fpage>. doi: <pub-id pub-id-type="doi">10.7554/eLife.62293</pub-id></citation>
</ref>
<ref id="ref115">
<label>115.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Kohonen</surname> <given-names>T</given-names></name>
</person-group>. <article-title>The self-organizing map</article-title>. <source>Proc IEEE</source>. (<year>1990</year>) <volume>78</volume>:<fpage>1464</fpage>&#x2013;<lpage>80</lpage>. doi: <pub-id pub-id-type="doi">10.1109/5.58325</pub-id></citation>
</ref>
<ref id="ref116">
<label>116.</label>
<citation citation-type="other"><person-group person-group-type="author"><name><surname>He</surname> <given-names>K.</given-names></name> <name><surname>Zhang</surname> <given-names>X.</given-names></name> <name><surname>Ren</surname> <given-names>S.</given-names></name> <name><surname>Sun</surname> <given-names>J.</given-names></name></person-group> Deep residual learning for image recognition. In 2016 IEEE conference on computer vision and pattern recognition (CVPR) (pp. 770&#x2013;778). IEEE; (<year>2016</year>).</citation>
</ref>
<ref id="ref117">
<label>117.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Cao</surname> <given-names>Y</given-names></name> <name><surname>Liu</surname> <given-names>W</given-names></name> <name><surname>Zhang</surname> <given-names>S</given-names></name> <name><surname>Xu</surname> <given-names>L</given-names></name> <name><surname>Zhu</surname> <given-names>B</given-names></name> <name><surname>Cui</surname> <given-names>H</given-names></name> <etal/></person-group>. <article-title>Detection and localization of myocardial infarction based on multi-scale resnet and attention mechanism</article-title>. <source>Front Physiol</source>. (<year>2022</year>) <volume>13</volume>:<fpage>783184</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fphys.2022.783184</pub-id>, PMID: <pub-id pub-id-type="pmid">35153827</pub-id></citation>
</ref>
<ref id="ref118">
<label>118.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Habijan</surname> <given-names>M</given-names></name> <name><surname>Gali&#x0107;</surname> <given-names>I</given-names></name> <name><surname>Leventi&#x0107;</surname> <given-names>H</given-names></name> <name><surname>Romi&#x0107;</surname> <given-names>K</given-names></name></person-group>. <article-title>Whole heart segmentation using 3d fm-pre-resnet encoder&#x2013;decoder based architecture with variational autoencoder regularization</article-title>. <source>Appl Sci</source>. (<year>2021</year>) <volume>11</volume>:<fpage>3912</fpage>. doi: <pub-id pub-id-type="doi">10.3390/app11093912</pub-id></citation>
</ref>
<ref id="ref119">
<label>119.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Harrer</surname> <given-names>S</given-names></name>
</person-group>. <article-title>Attention is not all you need: the complicated case of ethically using large language models in healthcare and medicine</article-title>. <source>EBioMedicine</source>. (<year>2023</year>) <volume>90</volume>:<fpage>104512</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.ebiom.2023.104512</pub-id>, PMID: <pub-id pub-id-type="pmid">36924620</pub-id></citation>
</ref>
<ref id="ref120">
<label>120.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Vaswani</surname> <given-names>A</given-names></name> <name><surname>Shazeer</surname> <given-names>N</given-names></name> <name><surname>Parmar</surname> <given-names>N</given-names></name> <name><surname>Uszkoreit</surname> <given-names>J</given-names></name> <name><surname>Jones</surname> <given-names>L</given-names></name> <name><surname>Gomez</surname> <given-names>AN</given-names></name> <etal/></person-group>. <article-title>Attention is all you need</article-title>. <source>Adv Neural Inf Proces Syst</source>. (<year>2017</year>) <volume>30</volume>. doi: <pub-id pub-id-type="doi">10.48550/arXiv.1706.03762</pub-id></citation>
</ref>
<ref id="ref121">
<label>121.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Li</surname> <given-names>C</given-names></name> <name><surname>Wang</surname> <given-names>Q</given-names></name> <name><surname>Liu</surname> <given-names>X</given-names></name> <name><surname>Hu</surname> <given-names>B</given-names></name></person-group>. <article-title>An attention-based CoT-ResNet with channel shuffle mechanism for classification of alzheimers disease levels</article-title>. <source>Frontiers in Aging Neuroscience.</source> (<year>2022</year>) <volume>14</volume>, <fpage>930584</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fnagi.2022.930584</pub-id></citation>
</ref>
<ref id="ref122">
<label>122.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Ma</surname> <given-names>S</given-names></name> <name><surname>Wang</surname> <given-names>H</given-names></name> <name><surname>Ma</surname> <given-names>L</given-names></name> <name><surname>Wang</surname> <given-names>L</given-names></name> <name><surname>Wang</surname> <given-names>W</given-names></name> <name><surname>Huang</surname> <given-names>S</given-names></name> <etal/></person-group>. <article-title>The era of 1-bit LLMs: all large language models are in 1.58 bits</article-title>. <source>arXiv [Preprint]</source>. (<year>2024</year>). doi: <pub-id pub-id-type="doi">10.48550/arXiv.2402.17764</pub-id></citation>
</ref>
<ref id="ref123">
<label>123.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Truong-Quoc</surname> <given-names>C</given-names></name> <name><surname>Lee</surname> <given-names>JY</given-names></name> <name><surname>Kim</surname> <given-names>KS</given-names></name> <name><surname>Kim</surname> <given-names>DN</given-names></name></person-group>. <article-title>Prediction of DNA origami shape using graph neural network</article-title>. <source>Nat Mater</source>. (<year>2024</year>):<fpage>1</fpage>&#x2013;<lpage>9</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41563-024-01846-8</pub-id>, PMID: <pub-id pub-id-type="pmid">38486095</pub-id></citation>
</ref>
<ref id="ref124">
<label>124.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>In&#x00E1;cio</surname> <given-names>MHDA</given-names></name> <name><surname>Shah</surname> <given-names>M</given-names></name> <name><surname>Jafari</surname> <given-names>M</given-names></name> <name><surname>Shehata</surname> <given-names>N</given-names></name> <name><surname>Meng</surname> <given-names>Q</given-names></name> <name><surname>Bai</surname> <given-names>W</given-names></name> <etal/></person-group>. <article-title>Cardiac age prediction using graph neural networks</article-title>. <source>MedRxiv [Preprint]</source>. (<year>2023</year>). doi: <pub-id pub-id-type="doi">10.1101/2023.04.19.23287590</pub-id></citation>
</ref>
<ref id="ref125">
<label>125.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Gao</surname> <given-names>J</given-names></name> <name><surname>Liu</surname> <given-names>J</given-names></name> <name><surname>Xu</surname> <given-names>Y</given-names></name> <name><surname>Peng</surname> <given-names>D</given-names></name> <name><surname>Wang</surname> <given-names>Z</given-names></name></person-group>. <article-title>Brain age prediction using the graph neural network based on resting-state functional MRI in Alzheimer's disease</article-title>. <source>Front Neurosci</source>. (<year>2023</year>) <volume>17</volume>:<fpage>1222751</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fnins.2023.1222751</pub-id>, PMID: <pub-id pub-id-type="pmid">37457008</pub-id></citation>
</ref>
<ref id="ref126">
<label>126.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Papillon</surname> <given-names>M</given-names></name> <name><surname>Sanborn</surname> <given-names>S</given-names></name> <name><surname>Hajij</surname> <given-names>M</given-names></name> <name><surname>Miolane</surname> <given-names>N</given-names></name></person-group>. <article-title>Architectures of topological deep learning: A survey on topological neural networks</article-title>. <source>arXiv [preprint]</source>. (<year>2023</year>). doi: <pub-id pub-id-type="doi">10.48550/arXiv.2304.10031</pub-id></citation>
</ref>
<ref id="ref127">
<label>127.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Shah</surname> <given-names>M</given-names></name> <name><surname>de A In&#x00E1;cio</surname> <given-names>MH</given-names></name> <name><surname>Lu</surname> <given-names>C</given-names></name> <name><surname>Schiratti</surname> <given-names>PR</given-names></name> <name><surname>Zheng</surname> <given-names>SL</given-names></name> <name><surname>Clement</surname> <given-names>A</given-names></name> <etal/></person-group>. <article-title>Environmental and genetic predictors of human cardiovascular ageing</article-title>. <source>Nat Commun</source>. (<year>2023</year>) <volume>14</volume>:<fpage>4941</fpage>. doi: <pub-id pub-id-type="doi">10.1038/s41467-023-40566-6</pub-id>, PMID: <pub-id pub-id-type="pmid">37604819</pub-id></citation>
</ref>
<ref id="ref128">
<label>128.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Huang</surname> <given-names>Y</given-names></name> <name><surname>Xu</surname> <given-names>J</given-names></name> <name><surname>Zhou</surname> <given-names>Y</given-names></name> <name><surname>Tong</surname> <given-names>T</given-names></name> <name><surname>Zhuang</surname> <given-names>X</given-names></name><collab id="coll4">Alzheimer&#x2019;s Disease Neuroimaging Initiative (ADNI)</collab></person-group>. <article-title>Diagnosis of Alzheimer&#x2019;s disease via multi-modality 3D convolutional neural network</article-title>. <source>Front Neurosci</source>. (<year>2019</year>) <volume>13</volume>:<fpage>509</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fnins.2019.00509</pub-id>, PMID: <pub-id pub-id-type="pmid">31213967</pub-id></citation>
</ref>
<ref id="ref129">
<label>129.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Acosta</surname> <given-names>JN</given-names></name> <name><surname>Falcone</surname> <given-names>GJ</given-names></name> <name><surname>Rajpurkar</surname> <given-names>P</given-names></name> <name><surname>Topol</surname> <given-names>EJ</given-names></name></person-group>. <article-title>Multimodal biomedical AI</article-title>. <source>Nat Med</source>. (<year>2022</year>) <volume>28</volume>:<fpage>1773</fpage>&#x2013;<lpage>84</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41591-022-01981-2</pub-id></citation>
</ref>
<ref id="ref130">
<label>130.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Casella</surname> <given-names>B</given-names></name> <name><surname>Riviera</surname> <given-names>W</given-names></name> <name><surname>Aldinucci</surname> <given-names>M</given-names></name> <name><surname>Menegaz</surname> <given-names>G</given-names></name></person-group>. <article-title>MERGE: A model for multi-input biomedical federated learning</article-title>. <source>Patterns</source>. (<year>2023</year>) <volume>4</volume>:<fpage>100856</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.patter.2023.100856</pub-id>, PMID: <pub-id pub-id-type="pmid">38035188</pub-id></citation>
</ref>
<ref id="ref131">
<label>131.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Kalra</surname> <given-names>S</given-names></name> <name><surname>Wen</surname> <given-names>J</given-names></name> <name><surname>Cresswell</surname> <given-names>JC</given-names></name> <name><surname>Volkovs</surname> <given-names>M</given-names></name> <name><surname>Tizhoosh</surname> <given-names>HR</given-names></name></person-group>. <article-title>Decentralized federated learning through proxy model sharing</article-title>. <source>Nat Commun</source>. (<year>2023</year>) <volume>14</volume>:<volume>2899</volume>. doi: <pub-id pub-id-type="doi">10.1038/s41467-023-38569-4</pub-id>, PMID: <pub-id pub-id-type="pmid">37217476</pub-id></citation>
</ref>
<ref id="ref132">
<label>132.</label>
<citation citation-type="other"><person-group person-group-type="author"><name><surname>Zhang</surname> <given-names>Y.</given-names></name> <name><surname>Xiang</surname> <given-names>T.</given-names></name> <name><surname>Hospedales</surname> <given-names>T. M.</given-names></name> <name><surname>Lu</surname> <given-names>H.</given-names></name></person-group> (<year>2018</year>). Deep mutual learning. In Proc IEEE Conference on Computer Vision and Pattern Recognition, 4320&#x2013;4328. doi: <pub-id pub-id-type="doi">10.1109/CVPR.2018.00454</pub-id></citation>
</ref>
<ref id="ref133">
<label>133.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Huang</surname> <given-names>S</given-names></name>
</person-group>. <article-title>The molecular and mathematical basis of Waddington's epigenetic landscape: a framework for post-Darwinian biology?</article-title> <source>BioEssays</source>. (<year>2012</year>) <volume>34</volume>:<fpage>149</fpage>&#x2013;<lpage>57</lpage>. doi: <pub-id pub-id-type="doi">10.1002/bies.201100031</pub-id></citation>
</ref>
<ref id="ref134">
<label>134.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Mc Auley</surname> <given-names>MT</given-names></name> <name><surname>Mooney</surname> <given-names>KM</given-names></name></person-group>. <article-title>Computational systems biology for aging research</article-title>. <source>Interdiscip Top Gerontol</source>. (<year>2015</year>) <volume>40</volume>:<fpage>35</fpage>&#x2013;<lpage>48</lpage>. doi: <pub-id pub-id-type="doi">10.1159/000364928</pub-id></citation>
</ref>
<ref id="ref135">
<label>135.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Tyson</surname> <given-names>JJ</given-names></name> <name><surname>Novak</surname> <given-names>B</given-names></name></person-group>. <article-title>A dynamical paradigm for molecular cell biology</article-title>. <source>Trends Cell Biol</source>. (<year>2020</year>) <volume>30</volume>:<fpage>504</fpage>&#x2013;<lpage>15</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.tcb.2020.04.002</pub-id>, PMID: <pub-id pub-id-type="pmid">32362451</pub-id></citation>
</ref>
<ref id="ref136">
<label>136.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Ni</surname> <given-names>SH</given-names></name> <name><surname>Xu</surname> <given-names>JD</given-names></name> <name><surname>Sun</surname> <given-names>SN</given-names></name> <name><surname>Li</surname> <given-names>Y</given-names></name> <name><surname>Zhou</surname> <given-names>Z</given-names></name> <name><surname>Li</surname> <given-names>H</given-names></name> <etal/></person-group>. <article-title>Single-cell transcriptomic analyses of cardiac immune cells reveal that Rel-driven CD72-positive macrophages induce cardiomyocyte injury</article-title>. <source>Cardiovasc Res</source>. (<year>2022</year>) <volume>118</volume>:<fpage>1303</fpage>&#x2013;<lpage>20</lpage>. doi: <pub-id pub-id-type="doi">10.1093/cvr/cvab193</pub-id>, PMID: <pub-id pub-id-type="pmid">34100920</pub-id></citation>
</ref>
<ref id="ref137">
<label>137.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Bergen</surname> <given-names>V</given-names></name> <name><surname>Soldatov</surname> <given-names>RA</given-names></name> <name><surname>Kharchenko</surname> <given-names>PV</given-names></name> <name><surname>Theis</surname> <given-names>FJ</given-names></name></person-group>. <article-title>RNA velocity&#x2014;current challenges and future perspectives</article-title>. <source>Mol Syst Biol</source>. (<year>2021</year>) <volume>17</volume>:<fpage>e10282</fpage>. doi: <pub-id pub-id-type="doi">10.15252/msb.202110282</pub-id>, PMID: <pub-id pub-id-type="pmid">34435732</pub-id></citation>
</ref>
<ref id="ref138">
<label>138.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Sinha</surname> <given-names>KK</given-names></name> <name><surname>Bilokapic</surname> <given-names>S</given-names></name> <name><surname>Du</surname> <given-names>Y</given-names></name> <name><surname>Malik</surname> <given-names>D</given-names></name> <name><surname>Halic</surname> <given-names>M</given-names></name></person-group>. <article-title>Histone modifications regulate pioneer transcription factor cooperativity</article-title>. <source>Nature</source>. (<year>2023</year>) <volume>619</volume>:<fpage>378</fpage>&#x2013;<lpage>84</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41586-023-06112-6</pub-id>, PMID: <pub-id pub-id-type="pmid">37225990</pub-id></citation>
</ref>
<ref id="ref139">
<label>139.</label>
<citation citation-type="journal"><person-group person-group-type="author"><name><surname>Barbuti</surname> <given-names>R</given-names></name> <name><surname>Gori</surname> <given-names>R</given-names></name> <name><surname>Milazzo</surname> <given-names>P</given-names></name> <name><surname>Nasti</surname> <given-names>L</given-names></name></person-group>. <article-title>A survey of gene regulatory networks modelling methods: from differential equations, to Boolean and qualitative bioinspired models</article-title>. <source>J Membr Comput</source>. (<year>2020</year>) <volume>2</volume>:<fpage>207</fpage>&#x2013;<lpage>26</lpage>. doi: <pub-id pub-id-type="doi">10.1007/s41965-020-00046-y</pub-id></citation>
</ref>
</ref-list>
</back>
</article>