EpiRL: A Reinforcement Learning Agent to Facilitate Epistasis Detection

TL;DR

This paper introduces EpiRL, a reinforcement learning approach that models gene-gene interaction detection as a Markov Decision Process, enabling efficient identification of significant epistasis in genetic data.

Contribution

It presents a novel RL-based framework for epistasis detection, addressing computational challenges of previous methods by framing the problem as a one-step MDP.

Findings

Successfully identifies highly interacted genes

Outperforms traditional methods in efficiency

Provides a new RL-based paradigm for genetic interaction analysis

Abstract

Epistasis (gene-gene interaction) is crucial to predicting genetic disease. Our work tackles the computational challenges faced by previous works in epistasis detection by modeling it as a one-step Markov Decision Process where the state is genome data, the actions are the interacted genes, and the reward is an interaction measurement for the selected actions. A reinforcement learning agent using policy gradient method then learns to discover a set of highly interacted genes.