Determines methylation state of DNA CpG dinucleotides using features inferred from three-dimensional genome topology (based on Hi-C) and DNA sequence patterns. DeepMethyl is a program that includes two stages: a supervised fine-tuning stage using labeled data (data with known target values) and an unsupervised pre-training stage using unlabeled training data. This tool can be used with sequential features generated within a window of the genome or features generated from the three-dimensional topology of a genome.
School of Computing, University of Southern Mississippi, Hattiesburg, MS, USA; Department of Computer Science and Christopher S. Bond Life Sciences Center, University of Missouri, Columbia, MO, USA; Department of Biochemistry and Molecular Biology, Georgia Regents University, Augusta, GA, USA; Department of Pharmacology and Toxicology, University of Mississippi Medical Center, North State Street, Jackson, MS, USA
DeepMethyl funding source(s)
Supported by the startup grant and a seed grant from National Science Foundation Mississippi EPSCoR (sub-award from Mississippi State University).