Enhancer Prediction and Feature Ranking of Enhancer Marks

Several epigenomic marks are available for predicting enhancers, but the integration of these marks continues to be a challenge. Three statistical models – a hidden Markov model, a support vector machine, and logistic regression – were implemented to integrate the data and better understand these enhancer marks, both in terms of their correlation with each other and their power to predict enhancers. The combined results provide further insight into the interrelations between enhancer marks and demonstrate that certain enhancer marks are more predictive