Machine learning for small interfering RNAs: a concise review of recent developments

The advent of machine learning and its subsequent integration into small interfering RNA (siRNA) research heralds a new epoch in the field of RNA interference (RNAi). This review emphasizes the urgency and relevance of assimilating the plethora of contributions and advancements in this domain, particularly focusing on the period of 2019–2023. Given the rapid progression of deep learning technologies, our synthesis of recent research is paramount to staying apprised of the state-of-the-art methods being utilized. It not only offers a comprehensive insight into the confluence of machine learning and siRNA but also serves as a beacon, guiding future explorations in this intersectional research field. Our rigorous examination of studies promises a discerning perspective on the contemporary landscape of machine learning applications in siRNA design and function. This review is an effort to foster further discourse and propel academic inquiry in this multifaceted domain.

1 Introduction

Enveloped within the expansive discipline of RNA interference (RNAi) (Wilson and Doudna, 2013; Mansoori et al., 2014; Rosa et al., 2018), the integration of machine learning strategies (Walia et al., 2012; Liu, 2019; Petegrosso et al., 2020) in the design and analysis of small interfering RNAs (siRNAs) marks a significant step in the advancement of this field. siRNAs, as vital components of the RNAi pathway, play an indispensable role in post-transcriptional gene silencing, influencing various genetic processes and, by extension, the potential for therapeutic interventions (Reynolds et al., 2004; Kanasty et al., 2013; Resnier et al., 2013; Dana et al., 2017; Tatiparti et al., 2017; Hu et al., 2020). Our review ventures into this rapidly evolving field, providing a detailed narrative of the seminal research contributions that blend the potent capabilities of machine learning with the inherent complexities of siRNA design and function.

Machine learning employs algorithms that improve automatically through experience (Jordan and Mitchell, 2015; Waring et al., 2020; Greener et al., 2022). It is employed across a myriad of applications, ranging from recommendation systems (Batmaz et al., 2019) to autonomous driving Bachute and Subhedar (2021), and now, increasingly in life sciences (Roscher et al., 2020). The unprecedented pace of machine learning advancements accentuates the need for an in-depth review of the most recent studies, ensuring that researchers and practitioners are abreast with state-of-the-art applications in the field.

It is this synthesis of machine learning and siRNA, an emergent and vital topic, that captures our academic interest. As the landscape of machine learning continues to diversify and mature, and siRNA’s influence in genetic research and therapeutic innovation becomes more profound, our review serves as a catalyst for fostering academic dialogue and nurturing exploratory research. Herein, we have carefully investigated studies published between 2019 and 2023 through Web of Science (WoS) using keywords of machine learning and siRNAs. The utilization of the WoS platform stems from its comprehensive incorporation of solely peer-reviewed journal articles of high quality. The specific timeframe of 2019–2023 was chosen not solely based on the quantity of research produced, but also due to the significant advancements in machine learning techniques, particularly deep learning methodologies, during this period.

This academic landscape is mirrored by a discernible gap in research studies focusing on machine learning applications for siRNAs, as shown in Figure 1. The result through the Web of Science database reveals a significantly lesser number of publications on machine learning and siRNAs as compared to other RNA-related topics, such as CRISPR and RNA-binding proteins (RBPs). This indicates an under-explored niche in the application of machine learning methods for siRNA analysis and design. The reviewed studies are summarized in Table 1, weaving together a comprehensive and up-to-date review of this intersectional field.

FIGURE 1

FIGURE 1. Comparison of the number of research papers retrieved from the Web of Science (WoS) database when searched with the keyword “Machine Learning.”

TABLE 1

TABLE 1. Overview of recent studies using machine learning methods in siRNA.

In the course of our comprehensive review of recent developments in the application of machine learning for siRNAs, it is observed that an intriguing distribution of machine learning models were implemented in the examined studies. The predominant model of choice was the Neural Network (NN), utilized in a total of 10 studies, illustrating a preference for its ability to model complex non-linear relationships and its inherent aptitude for handling high-dimensional data typical in siRNA research. The Support Vector Machine (SVM) model was adopted in five studies, reflecting its well-regarded robustness and efficacy in dealing with both linear and non-linear classification problems. Meanwhile, the Random Forest (RF) model was employed in three studies, underscoring its suitability for managing multi-dimensional datasets with its ensemble-based approach and inherent feature selection mechanism. Lastly, the Partial Least Squares (PLS) model was applied in two studies, illustrating its utilization in situations where predictors are highly correlated, a common occurrence in biological data. The majority of studies employed correlation and least squares as performance metrics in their analyses. Notably, some studies deployed multiple models, acknowledging the unique strengths of each model and adopting a more holistic, hybrid approach to tackle the multifaceted complexities inherent in siRNA design and analysis.

2 Machine learning methods for small interfering RNAs

2.1 Predictions of siRNA efficacy and off-target effects

Machine learning has emerged as a crucial tool in the field of siRNA research, facilitating nuanced investigations into the complex dynamics of siRNA. Two particularly salient areas of exploration are the prediction of siRNA efficacy and the elucidation of off-target effects. A comparative analysis of the studies by La Rosa et al. (2022) and Metwally et al. (2022) reveals distinct approaches to the application of machine learning methodologies for predicting siRNA efficacy.

La Rosa et al. (2022) implemented a novel Graph Neural Network (GNN) for evaluating siRNA-mRNA interaction networks, with an aim to predict siRNA efficacy. This approach marked a significant stride in the research area as GNNs, which outperformed conventional machine learning algorithms, were introduced for the first time in this context. Their method proved successful with a notable Pearson correlation coefficient of approximately 73.6%, representing the siRNA’s ability to bind and silence a gene target effectively.

On the other hand, Metwally et al. (2022) took a different approach by constructing a machine learning model for in silico prediction of siRNA ionizable-lipid nanoparticles’ in vivo efficacy. The authors adopted an array of machine learning techniques, including Artificial Neural Networks (ANNs) and SVM, for Quantitative Structure-Activity Relationship (QSAR) modeling, signifying a broader perspective of machine learning implementation. Notably, their model successfully predicted the siRNA dose, with ANNs delivering the most robust performance.

Exploring off-target effects and RNA production regulation, Müllerr et al. (2021) utilized supervised machine learning to adjust for variables that indirectly influence global RNA production in HeLa cells. This work provides an extensive dataset that paves the way for future exploration into global RNA metabolism regulation and its correlation with cellular states. Conversely, Kobayashi et al. (2022) focused on the off-target effects of siRNA, demonstrating that such effects are influenced by the base-pairing stability of two distinct regions with contrasting effects. Their thorough examination of siRNA’s subregions via an array of machine learning techniques established an important correlation between thermodynamic properties and off-target influence, thereby enhancing our understanding of siRNA’s off-target effects.

Both the methodological approaches, i.e., GNN and ANNs/SVM, show significant potential in siRNA research, albeit in different contexts. While GNN exhibited a superior ability in siRNA-mRNA interaction analysis, the versatility of ANNs/SVM was beneficial in predicting in vivo efficacy of siRNA nanoparticles. Further, machine learning proved to be instrumental in understanding off-target effects and RNA production regulation, demonstrating the versatility and potential of these techniques in decoding the complexities of siRNA.

2.2 Unveiling cellular processes involving siRNA

The integration of high-throughput screening in siRNA studies has ushered in a new era of comprehensive insights into cellular behavior and mechanisms under the influence of genetic modification. These following articles underscore this trend, showcasing how machine learning techniques are employed to delve deeper into proteome regulation, cellular delivery, and phenotype expression.

Scott et al. (2020) and Kustatscher et al. (2023) provide perspectives on the use of machine learning in high-content screening and proteome regulation. Both studies underscore the effectiveness of machine learning in uncovering complex biological systems. Scott et al. (2020) used multiparameter principal component analysis and an unbiased, parameter-agnostic machine learning approach to uncover genes and pathways that regulate mitochondrial clearance. This approach allowed the exploration of siRNA-based screening data in detail and led to the identification of modulators of parkin recruitment to mitochondria.

On the other hand, Kustatscher et al. (2023) took a higher-level perspective, using machine learning to analyze proteomics data and discover co-regulation modules, termed “progulons”. This approach offered a robust framework for studying the human proteome, identifying 31 progulons that constitute core cellular functions. Supervised machine learning, in this case, not only facilitated data processing but also uncovered new replication factors. The comparison underscores the flexibility and utility of machine learning applications in different research settings and objectives, both achieving significant findings in their respective domains.

Two recent studies provided insights into the usage of machine learning for precision delivery and phenotype assessment in siRNA studies, respectively. Patino et al. (2022) demonstrated the use of a live-cell analysis device (LCAD) coupled with deep learning to perform localized electroporation-induced membrane permeabilization, allowing precise siRNA delivery and content extraction from live cells. The combination of deep learning with LCAD technology represents a synergistic integration of novel hardware and advanced analytical tools, suggesting new opportunities for precise genetic interventions and real-time cellular response monitoring. Eismann et al. (2020) introduced an automated screening workflow using light-sheet microscopy to evaluate mitotic phenotypes in 3D cell cultures following siRNA knockdown. They employed a convolutional neural network (CNN) for phenotype classification, achieving high-throughput screening with high spatiotemporal resolution. This methodology enables a precise assessment of mitotic phenotypes in an automated, high-throughput manner, highlighting the power of deep learning in image processing and phenotype recognition.

These studies underscore the significant potential of machine learning, from neural networks to principal component analysis, in advancing siRNA research. Whether it is the identification of new cellular pathways, the precise delivery of functional molecules, or the high-throughput screening of phenotypes, machine learning methodologies emerge as vital tools, showcasing the increasing intersection between computational and biological sciences.

2.3 Elucidating the role of siRNA in diseases

Biomedical research has witnessed the transformative potential of machine learning, catalyzing breakthroughs in disease diagnosis and prognosis. Two notable applications of these technologies involve the identification of prognostically significant genes in cancer and the discovery of diagnostic gene biomarkers. In these contexts, the incorporation of deep learning methods has yielded significant insights.

The study by Kukita et al. (2022) demonstrated the use of machine learning in combination with siRNA, chromatin immunoprecipitation sequencing, and RNA sequencing for pinpointing prognostically significant genes, focusing on endometrial cancer. The researchers identified that the histone methyltransferase SETD8 regulates gene expression via H4K20 methylation and the p53 signaling pathway. Interestingly, they observed that suppressing SETD8, through siRNA or a selective inhibitor, could potentially inhibit cell proliferation and instigate apoptosis in endometrial cancer cells. This example of machine learning implementation showcases the potency of these methods in generating meaningful and impactful discoveries in cancer research.

On the other hand, Sun et al. (2022) used machine learning to identify diagnostic gene markers associated with immune infiltration in patients with renal fibrosis. They integrated Support Vector Machine Recursive Feature Elimination (SVM-RFE) and Least Absolute Shrinkage and Selection Operator (LASSO) regression models to achieve this. Their study identified nine key genes, with the knockdown of ISG20 via siRNA significantly inhibiting renal fibrosis progression in vitro. This study is a compelling example of how machine learning can drive novel insights in diagnostic biomarker discovery and influence therapeutic strategies.

Both studies demonstrate the profound potential of machine learning in the exploration of disease genetics, either in a prognostic or diagnostic capacity. However, the approaches vary in their specificity. The approach by Kukita et al. (2022) primarily focused on the downstream effects of a specific gene (SETD8), whereas the method by Sun et al. (2022) was more general, analyzing a broader set of potential markers. Despite these differences, both studies effectively incorporated machine learning to inform and enrich our understanding of disease biology.

2.4 siRNA delivery and drug discovery

The convergence of deep learning and machine learning techniques is propelling siRNA research, particularly in designing efficient delivery systems and accelerating the drug discovery process. As exemplified in the study by Kavya et al. (2022), an Artificial Neural Network (ANN) model was utilized to predict the release behavior of drugs and genes from a curcumin-loaded polymer synthesized in supercritical CO₂, encapsulating both curcumin and Bcl₂ siRNA. The promising results obtained underscore the potential of deep learning models in predicting siRNA delivery and release patterns, thus potentially revolutionizing the design of more effective siRNA delivery systems.

On the other hand, Kuthuru et al. (2019) leveraged machine learning methodologies to predict drug-kinase-target interactions from a high-content analysis data from an siRNA human kinome screen. They developed two types of kinase descriptors and applied machine learning models to predict these interactions, with the top model achieving an area under the ROC curve of 0.86. This clearly indicates the potential of machine learning in expediting the process of drug discovery by accurately predicting drug-target interactions.

2.5 Other emerging topics in siRNA research

The rise of machine learning techniques has brought about a significant paradigm shift in siRNA research. Recent investigations have successfully harnessed both traditional machine learning and deep learning approaches to address challenges and answer pivotal questions in the field. This subsection provides the recent trend of emerging machine learning methodologies in siRNA studies, focusing on their unique applications, effectiveness, and particular roles in advancing siRNA research.

In relation to nanopore technology, Penguin (Hassan et al., 2022) and Sequoia (Koonchanok et al., 2023) have emerged as significant tools that leverage machine learning for direct RNA sequencing data analysis. Penguin is designed to identify pseudouridine sites in RNA, employing machine learning models such as SVM, RF, and NN to process the raw signal generated by Oxford Nanopore sequencing. On the other hand, Sequoia provides a comprehensive framework for visual analysis of RNA modifications from nanopore sequencing data, enabling users to interactively analyze and cluster signals based on electric-current similarities. In structural profiling of low molecular weight RNAs, Wang et al. (2021) proposed a novel machine learning algorithm to augment nanopore trapping/translocation. This algorithm transformed raw event characteristics into interpretable data, with an impressive accuracy of approximately 93.4%. Importantly, the algorithm was able to distinguish between various RNA types, demonstrating its potential for future siRNA studies. On a different note, He et al. (2019) applied a deep learning-based approach for predicting virus-derived small interfering RNAs (vsiRNAs) in plants. Their deep Convolutional Neural Network (CNN) model, PVsiRNAPred, trained on vsiRNA sequences, demonstrated superior performance to five conventional machine learning classifiers, achieving an accuracy of 65.70%. Both studies utilized machine learning, but their focus and approach differed, reflecting the versatility of machine learning applications in siRNA research.

Nademi et al. (2021) employed machine learning for predicting the cellular uptake of hydrophobically modified Polyethylenimine (PEI)/siRNA nanoparticles in various cancer cell lines. Using three regression models, the study revealed that non-linear models, such as RF and Multilayer Perceptron (MLP), outperformed the Linear Regression model in predictive accuracy. The predictive performance of these non-linear models shows their potential in improving our understanding of siRNA nanoparticle uptake in cancer research.

Meanwhile, Persson Hoden et al. (2021) developed an R package, smartPARE, that utilizes a deep learning CNN for the identification of true mRNA cleavage sites. Applied to high-throughput datasets, smartPARE effectively identified true cleavage sites, providing crucial insights into the small RNA (sRNA) landscape in complex biological systems.

In conclusion, the application of machine learning, from traditional algorithms to deep learning methods, is proving vital in various aspects of siRNA research, including structural profiling, prediction of vsiRNAs, cellular uptake prediction, and identification of mRNA cleavage sites. Although the methods and applications vary, the overall advancement in the field signifies the transformative potential of machine learning in this area. The comparison further highlights the benefits of non-linear and deep learning models over traditional linear models in terms of predictive accuracy and versatility, leading to valuable discoveries in the field of siRNA research.

3 Discussions: Future perspectives and challenges

The confluence of machine learning and siRNA research has already brought to light numerous applications and technological advancements. From predicting siRNA efficacy and off-target effects, to uncovering cellular processes involving siRNA and elucidating the role of siRNA in diseases, the combination of these fields has opened up new avenues for scientific exploration and innovation. However, similar to all emerging fields, it comes with its own set of unique challenges and limitations.

The application of machine learning techniques in siRNA research is hampered by the lack of expansive, high-fidelity datasets. This is a common problem in many machine learning applications, but it is particularly pronounced in the field of biological research, where experimental data can be costly and time-consuming to generate. However, as techniques such as high-throughput screening continue to advance, the availability of high-quality datasets for siRNA research is expected to increase. Moreover, strategies such as transfer learning and data augmentation could be leveraged to overcome the scarcity of data and enhance the learning capacity of machine learning models. As an alternative approach, the high-content screening techniques proposed by Scott et al. (2020) can be used to address this limitation of high-fidelity datasets.

Another significant challenge is the inherent complexity of biological systems. The multitude of interacting factors and the non-linear nature of biological processes pose significant difficulties for the construction and optimization of machine learning models. However, novel machine learning methods, such as GNN and deep learning (La Rosa et al., 2022), have demonstrated promising results in dealing with such complexities. Future work should continue to explore and optimize these techniques for application in siRNA research.

Finally, the lack of interpretable machine learning models in this field is a crucial area that needs to be addressed (Murdoch et al., 2019). Despite the promising results achieved by complex models such as deep learning, their black-box nature poses a significant challenge for their broader acceptance and utilization. Therefore, the development and application of more transparent, interpretable models should be a priority for future research.

While challenges and obstacles lie ahead, the potential rewards of integrating machine learning and siRNA research are vast. It is our hope that this review will provide a useful roadmap for future researchers navigating this exciting and rapidly evolving field.

4 Conclusion

In conclusion, the fusion of machine learning and siRNA marks a promising Frontier in the realm of RNA interference. Although faced with challenges, such as the need for large, high-quality datasets and the intricate nature of biological systems, the continued development of advanced machine learning models and feature engineering techniques offers an optimistic outlook on the field’s future.

The rapidly evolving landscape of machine learning necessitates frequent and thorough investigation of recent studies, particularly when coupled with the emergent field of siRNA. This review has thus aimed to provide a comprehensive, timely exploration of these two intertwined fields, bridging the gap between computational advancements and biological complexities. It is our fervent hope that this work will serve as a foundation for future explorations and will inspire novel, cross-disciplinary endeavors.

Author contributions

ML contributed to conception and design of the study and wrote sections of the manuscript.

Funding

This work was supported by a research grant funded by Generative Artificial Intelligence System Inc. (GAIS). The funding sponsor had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.

Conflict of interest

ML has received research grants from Generative Artificial Intelligence System Inc.

Publisher’s note

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.

References

Bachute, M. R., and Subhedar, J. M. (2021). Autonomous driving architectures: Insights of machine learning and deep learning algorithms. Mach. Learn. Appl. 6, 100164. doi:10.1016/j.mlwa.2021.100164

CrossRef Full Text | Google Scholar

Batmaz, Z., Yurekli, A., Bilge, A., and Kaleli, C. (2019). A review on deep learning for recommender systems: Challenges and remedies. Artif. Intell. Rev. 52, 1–37. doi:10.1007/s10462-018-9654-y

CrossRef Full Text | Google Scholar

Dana, H., Chalbatani, G. M., Mahmoodzadeh, H., Karimloo, R., Rezaiean, O., Moradzadeh, A., et al. (2017). Molecular mechanisms and biological functions of sirna. Int. J. Biomed. Sci. IJBS 13, 48–57.

PubMed Abstract | Google Scholar

Eismann, B., Krieger, T. G., Beneke, J., Bulkescher, R., Adam, L., Erfle, H., et al. (2020). Automated 3d light-sheet screening with high spatiotemporal resolution reveals mitotic phenotypes. J. Cell. Sci. 133, jcs245043. doi:10.1242/jcs.245043

PubMed Abstract | CrossRef Full Text | Google Scholar

Greener, J. G., Kandathil, S. M., Moffat, L., and Jones, D. T. (2022). A guide to machine learning for biologists. Nat. Rev. Mol. Cell. Biol. 23, 40–55. doi:10.1038/s41580-021-00407-0

PubMed Abstract | CrossRef Full Text | Google Scholar

Hassan, D., Acevedo, D., Daulatabad, S. V., Mir, Q., and Janga, S. C. (2022). Penguin: A tool for predicting pseudouridine sites in direct rna nanopore sequencing data. Methods 203, 478–487. doi:10.1016/j.ymeth.2022.02.005

PubMed Abstract | CrossRef Full Text | Google Scholar

He, B., Huang, J., and Chen, H. (2019). Pvsirnapred: Prediction of plant exclusive virus-derived small interfering rnas by deep convolutional neural network. J. Bioinforma. Comput. Biol. 17, 1950039. doi:10.1142/S0219720019500392

CrossRef Full Text | Google Scholar

Hu, B., Zhong, L., Weng, Y., Peng, L., Huang, Y., Zhao, Y., et al. (2020). Therapeutic sirna: State of the art. Signal Transduct. Target. Ther. 5, 101. doi:10.1038/s41392-020-0207-x

PubMed Abstract | CrossRef Full Text | Google Scholar

Jordan, M. I., and Mitchell, T. M. (2015). Machine learning: Trends, perspectives, and prospects. Science 349, 255–260. doi:10.1126/science.aaa8415

PubMed Abstract | CrossRef Full Text | Google Scholar

Kanasty, R., Dorkin, J. R., Vegas, A., and Anderson, D. (2013). Delivery materials for sirna therapeutics. Nat. Mater. 12, 967–977. doi:10.1038/nmat3765

PubMed Abstract | CrossRef Full Text | Google Scholar

Kavya, K. V., Vargheese, S., Shukla, S., Khan, I., Dey, D. K., Bajpai, V. K., et al. (2022). A cationic amino acid polymer nanocarrier synthesized in supercritical co2 for co-delivery of drug and gene to cervical cancer cells. Col Surf B-Bio 216, 112584. doi:10.1016/j.colsurfb.2022.112584

CrossRef Full Text | Google Scholar

Kobayashi, Y., Tian, S., and Ui-Tei, K. (2022). The sirna off-target effect is determined by base-pairing stabilities of two different regions with opposite effects. Genes. 13, 319. doi:10.3390/genes13020319

PubMed Abstract | CrossRef Full Text | Google Scholar

Koonchanok, R., Daulatabad, S. V., Reda, K., and Janga, S. C. (2023). “Sequoia: A framework for visual analysis of rna modifications from direct rna sequencing data,” in Computational epigenomics and epitranscriptomics (New York, NY: Humana), 127–138.

CrossRef Full Text | Google Scholar

Kukita, A., Sone, K., Kaneko, S., Kawakami, E., Oki, S., Kojima, M., et al. (2022). The histone methyltransferase setd8 regulates the expression of tumor suppressor genes via h4k20 methylation and the p53 signaling pathway in endometrial cancer cells. Cancers 14, 5367. doi:10.3390/cancers14215367

PubMed Abstract | CrossRef Full Text | Google Scholar

Kustatscher, G., Hodl, M., Rullmann, E., Grabowski, P., Fiagbedzi, E., Groth, A., et al. (2023). Higher-order modular regulation of the human proteome. Mol. Syst. Biol. 19, e9503. doi:10.15252/msb.20209503

PubMed Abstract | CrossRef Full Text | Google Scholar

Kuthuru, S., Szafran, A. T., Stossi, F., Mancini, M. A., and Rao, A. (2019). Leveraging image-derived phenotypic measurements for drug-target interaction predictions. Cancer Inf. 18, 1176935119856595. doi:10.1177/1176935119856595

CrossRef Full Text | Google Scholar

La Rosa, M., Fiannaca, A., La Paglia, L., and Urso, A. (2022). A graph neural network approach for the analysis of sirna-target biological networks. Int. J. Mol. Sci. 23, 14211. doi:10.3390/ijms232214211

PubMed Abstract | CrossRef Full Text | Google Scholar

Liu, B. (2019). Bioseq-analysis: A platform for dna, rna and protein sequence analysis based on machine learning approaches. Briefings Bioinforma. 20, 1280–1294. doi:10.1093/bib/bbx165

CrossRef Full Text | Google Scholar

Mansoori, B., Shotorbani, S. S., and Baradaran, B. (2014). Rna interference and its role in cancer therapy. Adv. Pharm. Bull. 4, 313–321. doi:10.5681/apb.2014.046

PubMed Abstract | CrossRef Full Text | Google Scholar

Metwally, A. A., Nayel, A. A., and Hathout, R. M. (2022). In silico prediction of sirna ionizable-lipid nanoparticles in vivo efficacy: Machine learning modeling based on formulation and molecular descriptors. Front. Mol. Biosci. 9, 1042720. doi:10.3389/fmolb.2022.1042720

PubMed Abstract | CrossRef Full Text | Google Scholar

Müller, M., Avar, M., Heinzer, D., Emmenegger, M., Aguzzi, A., Pelkmans, L., et al. (2021). High content genome-wide sirna screen to investigate the coordination of cell size and rna production. Sci. Data 8, 162. doi:10.1038/s41597-021-00944-5

PubMed Abstract | CrossRef Full Text | Google Scholar

Murdoch, W. J., Singh, C., Kumbier, K., Abbasi-Asl, R., and Yu, B. (2019). Definitions, methods, and applications in interpretable machine learning. Proc. Natl. Acad. Sci. 116, 22071–22080. doi:10.1073/pnas.1900654116

PubMed Abstract | CrossRef Full Text | Google Scholar

Nademi, Y., Tang, T., and Uludag, H. (2021). Modeling uptake of polyethylenimine/short interfering rna nanoparticles in breast cancer cells using machine learning. Adv. Nanobiomed Res. 1, 2000106. doi:10.1002/anbr.202000106

CrossRef Full Text | Google Scholar

Patino, C. A., Mukherjee, P., Berns, E. J., Moully, E. H., Stan, L., Mrksich, M., et al. (2022). High-throughput microfluidics platform for intracellular delivery and sampling of biomolecules from live cells. ACS Nano 16, 7937–7946. doi:10.1021/acsnano.2c00698

PubMed Abstract | CrossRef Full Text | Google Scholar

Persson Hoden, K., Hu, X., Martinez, G., and Dixelius, C. (2021). smartpare: An r package for efficient identification of true mrna cleavage sites. Int. J. Mol. Sci. 22, 4267. doi:10.3390/ijms22084267

PubMed Abstract | CrossRef Full Text | Google Scholar

Petegrosso, R., Li, Z., and Kuang, R. (2020). Machine learning and statistical methods for clustering single-cell rna-sequencing data. Briefings Bioinforma. 21, 1209–1223. doi:10.1093/bib/bbz063

CrossRef Full Text | Google Scholar

Resnier, P., Montier, T., Mathieu, V., Benoit, J.-P., and Passirani, C. (2013). A review of the current status of sirna nanomedicines in the treatment of cancer. Biomaterials 34, 6429–6443. doi:10.1016/j.biomaterials.2013.04.060

PubMed Abstract | CrossRef Full Text | Google Scholar

Reynolds, A., Leake, D., Boese, Q., Scaringe, S., Marshall, W. S., and Khvorova, A. (2004). Rational sirna design for rna interference. Nat. Biotechnol. 22, 326–330. doi:10.1038/nbt936

PubMed Abstract | CrossRef Full Text | Google Scholar

Rosa, C., Kuo, Y.-W., Wuriyanghan, H., and Falk, B. W. (2018). Rna interference mechanisms and applications in plant pathology. Annu. Rev. phytopathology 56, 581–610. doi:10.1146/annurev-phyto-080417-050044

CrossRef Full Text | Google Scholar

Roscher, R., Bohn, B., Duarte, M. F., and Garcke, J. (2020). Explainable machine learning for scientific insights and discoveries. Ieee Access 8, 42200–42216. doi:10.1109/access.2020.2976199

CrossRef Full Text | Google Scholar

Scott, H. L., Buckner, N., Fernandez-Albert, F., Pedone, E., Postiglione, L., Shi, G., et al. (2020). A dual druggable genome-wide sirna and compound library screening approach identifies modulators of parkin recruitment to mitochondria. J. Biol. Chem. 295, 3285–3300. doi:10.1074/jbc.RA119.009699

PubMed Abstract | CrossRef Full Text | Google Scholar

Sun, Y.-C., Qiu, Z.-Z., Wen, F.-L., Yin, J.-Q., and Zhou, H. (2022). Revealing potential diagnostic gene biomarkers associated with immune infiltration in patients with renal fibrosis based on machine learning analysis. J. Immunol. Res. 2022, 3027200. doi:10.1155/2022/3027200

PubMed Abstract | CrossRef Full Text | Google Scholar

Tatiparti, K., Sau, S., Kashaw, S. K., and Iyer, A. K. (2017). Sirna delivery strategies: a comprehensive review of recent developments. Nanomaterials 7, 77. doi:10.3390/nano7040077

PubMed Abstract | CrossRef Full Text | Google Scholar

Walia, R. R., Caragea, C., Lewis, B. A., Towfic, F., Terribilini, M., El-Manzalawy, Y., et al. (2012). Protein-rna interface residue prediction using machine learning: An assessment of the state of the art. BMC Bioinforma. 13, 89–20. doi:10.1186/1471-2105-13-89

CrossRef Full Text | Google Scholar

Wang, Y., Guan, X., Zhang, S., Liu, Y., Wang, S., Fan, P., et al. (2021). Structural-profiling of low molecular weight rnas by nanopore trapping/translocation using mycobacterium smegmatis porin a. Nat. Commun. 12, 3368. doi:10.1038/s41467-021-23764-y

PubMed Abstract | CrossRef Full Text | Google Scholar

Waring, J., Lindvall, C., and Umeton, R. (2020). Automated machine learning: Review of the state-of-the-art and opportunities for healthcare. Artif. Intell. Med. 104, 101822. doi:10.1016/j.artmed.2020.101822

PubMed Abstract | CrossRef Full Text | Google Scholar

Wilson, R. C., and Doudna, J. A. (2013). Molecular mechanisms of rna interference. Annu. Rev. biophysics 42, 217–239. doi:10.1146/annurev-biophys-083012-130404

CrossRef Full Text | Google Scholar