EpiML (https://epiml.uncc.edu/) provides a web service for detecting main effect and epistatic effect in genomic data. EpiML currently supports EBEN (Empirical Bayesian Elastic Net), LASSO and ssLASSO (spike-and-slab LASSO) for machine learning based epistasis analysis. In addition to interative visualizations, EpiML provides the download of data, results, and anlytical code/containers.
An illustration of the EpiML workflow is shown as follows: Figure 1. The workflow of the EpiML web service. a. Our epistasis analysis server allows users to describe their job, upload data and select a machine learning method for analysis. b. We provide interactive tools for visualizing and downloading results. c. Customize analysis. All automatically generated Jupyter notebooks and pre-configured Docker containers can be downloaded, modified and rerun, allowing users to fully customize their analysis code on local computers.