This respository includes the source code for the DeFCoM supervised ML genomic footprint classification framework described in this publication. If you use this tool in your research, please include the following citation:
Quach, B., & Furey, T. S. (2017). DeFCoM: analysis and modeling of transcription factor binding sites using a motif-centric genomic footprinter. Bioinformatics, 33(7), 956-963.
Note: This framework has not been maintained since 2017 and was written in Python 2.7. Please reach out if you are interested in collaborations to refactor the codebase. The manuscript branch includes source files that correspond to the manuscript version of the tool.
For a general overview of genomic footprinting see this review by Sung et al.(2016). DeFCoM was designed to learn patterns for chromatin accessibility signals at transcription factor binding sites across the genome from DNaseI-seq or ATAC-seq data. These learned transcription factor binding patterns can then be used to predict binding sites in new DNaseI-seq or ATAC-seq samples. The following image provides an overview of the DeFCoM framework:
For detailed documentation see the Read the Docs page.