Introduction

Improse is a supervised machine learning approach to predict super-enhancers or constituents of super-enhancers for a list of candidate enhancers. Improse integrated diverse features including DNase I hypersensitivity (DNaseI), histone modifications (HMs), cofactors, transcription factors (TFs) and DNA sequence specific features.

Improse comes with six state-of-the-art machine learning models including Random Forest (RF), Support Vector Machines (SVM), K-Nearest Neighbor (kNN), AdaBoost (AB), Decision Tree (DT) and Naive Bayes (NB).

Random Forest is our optimal and default model but user can select any of the model and further test it using cross-validation, independent test data or to make predictions.