Introduction

For code and examples visit GitHub (Link).

GEP is machine learning based computational tool for genome-wide prediction of active enhancers in various cell-types and tissues. The rationale behind the tool is to learn exclusively from experimentally characterized active enhancers, those epigenomic patterns with discriminatory power by contrasting active enhancers with comparable background genomic regions. Importantly, the genomic background is structured in a balanced way representative of potential, non-enhancer regulatory elements and gene body regions, instead of randomly sampled genomic regions. The Random Forest based model is implemented in Python. The package contains accessary programs required for data processing.

• Quick-Start