Presentation Information
[SS15-03]Decoding the Role of Epigenetic Fluctuation in Stochastic Cell and Individual State Transitions
*Risa Karakida Kawaguchi1,2 (1. The University of Tokyo (Japan), 2. Kyoto University (Japan))
Keywords:
Machine Learning,Meta-analytic approach,Multi-omics data,Deep Learning
High-throughput sequencing technologies have long played a pivotal role in generating large-scale biological data, including at the single-cell level. However, large-scale omics analyses often face significant challenges due to batch effects, which can obscure true reproducible biological signals. In this talk, we introduce the utility of both meta-analytic and machine learning approaches for the analysis of bulk and single-cell omics data, by analyzing human, mouse, and nine-banded armadillo data. By training predictive models—particularly deep neural networks—we are able to identify regulatory mechanisms that govern distinct developmental stages and lineages. Furthermore, the variations that cannot be captured by these models may provide critical insights into how stochastic fluctuations, potentially through epigenetic modifications, influence cell fate decisions and individual phenotypes. Our results highlight the power of integrated computational approaches to unravel complex regulatory landscapes in diverse biological systems.