About Chromia

Chromia is supervised learning method that integrates epigenomic and genomic information to predict functional locis (e.g. promoters and enhancers) as well as TF binding sites.

Input Data

Training Data: Chromia is a supervised method using Hidden Markov models (HMMs), which can be trained on the histone modification data for known promoters and enhancers,so users should provide a promoter training file and an enhancer training file. The data files contain signature regions and the data format are here:

chr1    559755
chr2    149355835
chr5    134290695
chr5    134288135
chr17    21944875
chr1    91625495

Testing Data: Please upload multiple data files(usually each file contains the data of a specific ChIP-Seq mark). The input data format can be BED or WIG.


Chromia was proposed by Kyoung-Jae Won et al. The details can be found in the paper: Prediction of regulatory elements in mammalian genomes using chromatin signatures.