Bioinformatics. 2021 Jul 12;37(Supplement_1):i272-i279. doi: 10.1093/bioinformatics/btab272.
ABSTRACT
MOTIVATION: The high-throughput chromosome conformation capture (Hi-C) technique has enabled genome-wide mapping of chromatin interactions. However, high-resolution Hi-C data requires costly, deep sequencing; therefore, it has only been achieved for a limited number of cell types. Machine learning models based on neural networks have been developed as a remedy to this problem.
RESULTS: In this work, we propose a novel method, EnHiC, for predicting high-resolution Hi-C matrices from low-resolution input data based on a generative adversarial network (GAN) framework. Inspired by non-negative matrix factorization, our model fully exploits the unique properties of Hi-C matrices and extracts rank-1 features from multi-scale low-resolution matrices to enhance the resolution. Using three human Hi-C datasets, we demonstrated that EnHiC accurately and reliably enhanced the resolution of Hi-C matrices and outperformed other GAN-based models. Moreover, EnHiC-predicted high-resolution matrices facilitated the accurate detection of topologically associated domains and fine-scale chromatin interactions.
AVAILABILITY AND IMPLEMENTATION: EnHiC is publicly available at https://github.com/wmalab/EnHiC.
SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
PMID:34252966 | DOI:10.1093/bioinformatics/btab272