DeepChrome 2.0: Investigating and Improving Architectures, Visualizations, & Experiments
This work addresses gene regulation prediction in epigenetics, but it is incremental as it builds on prior DeepChrome research.
The study tackled predicting gene expression from histone modification signals by investigating and improving upon the DeepChrome model, finding that a 12-parameter linear network matches the predictive power of a 645k-parameter convolutional neural network and that the relationship is independent of cell type.
Histone modifications play a critical role in gene regulation. Consequently, predicting gene expression from histone modification signals is a highly motivated problem in epigenetics. We build upon the work of DeepChrome by Singh et al. (2016), who trained classifiers that map histone modification signals to gene expression. We present a novel visualization technique for providing insight into combinatorial relationships among histone modifications for gene regulation that uses a generative adversarial network to generate histone modification signals. We also explore and compare various architectural changes, with results suggesting that the 645k-parameter convolutional neural network from DeepChrome has the same predictive power as a 12-parameter linear network. Results from cross-cell prediction experiments, where the model is trained and tested on datasets of varying sizes, cell-types, and correlations, suggest the relationship between histone modification signals and gene expression is independent of cell type. We release our PyTorch re-implementation of DeepChrome on GitHub \footnote{\url{github.com/ssss1029/gene_expression_294}}.\parfillskip=0pt