GenoNet scores for human genome assembly GRCh38

Predicting the functional consequences of genetic variants in non-coding regions is a challenging problem. We propose here a semi-supervised approach, GenoNet, to jointly utilize experimentally confirmed regulatory variants (labeled variants), millions of unlabeled variants genome-wide, and more than a thousand cell/tissue type-specific epigenetic annotations to predict functional consequences of non-coding variants.

Format

The GenoNet scores are stored in the tab-delimited text files. 

Each row represents a genomic region with 131 columns. Please find the header line in "genonet.header.txt". 

The first four columns are chromosome, start coordinate, end coordinate, and a region ID named by positions. Please note that the coordinates are counted in the 0-based UCSC Genome Browser BED format. For example, the following region with a start position 10000 and an end position 10025 includes 25 base pairs within chr1:10001-10025.

chr1    10000    10025    chr1_10001_10025

Columns 5-131 are the predicted tissue-specific functional effects (GenoNet scores) for the 127 Roadmap tissues. Each column is named by the corresponding epigenome ID. This online spreadsheet includes the information about the 127 Roadmap tissues in detail.

Reference
Zihuai He, Linxi Liu, Kai Wang, Iuliana Ionita-Laza. A semi-supervised approach for predicting cell type/tissue specific functional consequences of non-coding variation using massively parallel reporter assays. Nature Communications, 2018.

Release

GRCh37 https://zenodo.org/record/3336209

GRCh38 liftover https://zenodo.org/record/6484230