About this Research Topic
The epigenome has been regarded as one of the most important regulators for genome on its functioning and downstream regulation, therefore, the maintenance of stable and normal functioning epigenome is quite crucial for living organisms. However, the epigenome changes over time, either naturally or triggered by various endogenous and exogenous factors like environment, disease state and infections. The alteration of the epigenome may be crucial intermediate between environments and biological phenotypes during development and pathogenesis. Therefore, developmental or pathological alterations can also be monitored using epigenome characteristics, which is exactly what epigenetics and epigenomics studies focus on.
Currently, studies on epigenetics and epigenomics mainly focused on the plasticity of epigenome during two major biological processes: development and pathogenesis, both of which are inextricably linked to the environment through the epigenetic modifications. From the research layers, current epigenomic and epigenetic studies can be further divided into multiple layers: 1) direct methylation on DNA molecules; 2) histone protein modification; 3) chromatin structure and 4) related noncoding RNAs. Integration all layers of epigenomics and epigenetics studies, the ultimate research goal in this field is to reveal the specific role of epigenome during the development and pathogenesis of human beings and explain the related biological mechanisms using typical epigenetics/epigenomics biomarkers.
Biologically, epigenetics and epigenomics describe complex interactions between environment and genomics, resulting in diverse modifications on histone and DNA molecules. With the development of detecting techniques (like microarray and Methyl-Seq), an explosive increase occurs in epigenetics and epigenomics data. To handle such massive complex data, machine learning models have been introduced in the analyses on data at this omics-level and contribute to the identification of potential disease/developmental events associated epigenetic biomarkers. However, several restrictions and challenges still remain in current epigenetics and epigenomics studies:
1) For most epigenomics studies, patients are hard to recruit (comparing to normal controls), lacking samples with diseases characteristics;
2) For each epigenome, epigenomic alterations with biological significance is highly imbalanced distributed across the genome, making it hard for us to detect;
3) Comparing to the sample number, methylation sites targeted by current probes are too many, forming a matrix with much more variables than samples. Larger datasets and pre-modeling features selection may be potential solutions for current restrictions on epigenetics and epigenomics studies.
In this research topic, we focused on the application of machine learning models on data at epigenomics and epigenetic levels to identify potential biomarkers for complex diseases/developmental events, including cancers. Study designs welcomed are listed as below:
1) Studies identifying new biomarkers for complex diseases/developmental events by either way including:
i. Studies based on relatively larger datasets or studies integrating epigenomics and epigenetics raw data from multiple previous publications to identify new biomarkers.
ii. Studies presenting effective pre-modelling feature selection methods to reduce the complexity of epigenomics and epigenetics datasets for further biomarker identification.
2) Studies connecting epigenomics data with other omics data, revealing the intermediate role of epigenomics and epigenetics biomarkers among environmental factors, multi-omics factors and diseases/developmental events.
3) Studies integrating new machine learning models/frameworks to identify new epigenomics and epigenetics biomarkers for complex diseases/developmental events.
Here, we encourage submissions of original research and review articles about the application of machine learning models to identify key epigenetic and epigenomic biomarkers for complex diseases or special developmental events. The applications of machine learning model to overcome current challenges and restrictions in epigenetic and epigenomics studies are all welcome, which include but not limited to new algorithm development, new biomarker identification, exploration of disease pathogenesis at epigenomics/epigenetics level, integrated analyses on multi-omics levels (including epigenomics level) and drug target prediction.
Declared competing interests: Michael Liebman is the co-founder of IPQ Analytics. The analytics platform (patent pending) is an ontology-based technology to extend knowledge dimensions spanning the clinical, molecular and commercial domains using natural language queries.
Keywords: Machine Learning, Epigenomics and Epigenetics, Biomarkers, Complex Diseases, Developmental Events
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.