henhaobio
Sign in

bioRxivGenomicsPreprint

mV2G: a multiomic atlas for tissue-specific variant-to-gene prioritization

Dean Y ZhangiD, Hufeng ZhouiD, Maya U. ShethiD, Andreas R. GschwindiD, J EngreitziD, Xihong LiniD, Hongbo LiuiD

PDFDOIData

In-depth summary

Claude Opus 5.5 · xhigh effort · 7 Nov 2026 · from the full text of version 1 (20 Aug 2026) · CC BY 4.0

A web atlas counts evidence from 24 types for 188,634,118 variant-gene pairs across 50 human tissues; 1,530,420 pairs pass a high-confidence cutoff, 87% of them tissue-specific.

Most GWAS-associated variants are noncoding, and the evidence linking them to target genes is spread across molecular quantitative trait locus (xQTL), enhancer-gene, chromatin-contact and CRISPR resources, while existing combined scores (cS2G, Open Targets, GeneHancer) lack tissue context. mV2G extends the authors' earlier kidney multiomic strategy to 50 tissues, scoring each variant-gene pair separately per tissue and showing the underlying evidence on a searchable website.

Key results

  1. ShownThe atlas holds 188,634,118 variant-gene pairs that have at least one supporting evidence type in at least one of 50 tissues. These pairs cover 13,618,039 variants and 69,521 genes.

    Built from 21 tissue-specific evidence types plus 3 tissue-agnostic strategies, applied to a reference of 425,576,568 candidate pairs within 1 Mbp. The genes include 18,403 protein-coding genes and 51,118 noncoding RNAs. Per-tissue counts of supported pairs ranged from 8,952,481 (brain amygdala) to 133,370,096 (whole blood). · Annotations of variants, genes, and…; V2G evidence integration for high-confidence V2G…; Overview of the mV2G atlas

  2. SuggestedRaising the threshold on the evidence count shrinks the prioritized set but raises the share of variants that point to a single gene. A score of at least 5 was adopted as the high-confidence cutoff.

    At a score of 5 or more, each of the 50 tissues had more than 80% of its prioritized variants linked to only one gene. The tissue curves differed in level, which the authors attribute to differences in the amount and composition of evidence available per tissue. The single-gene fraction is a proxy for confidence, not a validation against known causal genes. · V2G evidence integration for high-confidence V2G…; Prioritization of high-confidence V2G…

  3. ShownThe high-confidence set (score of 5 or more in any tissue) contains 1,530,420 pairs, involving 1,131,316 variants and 18,426 genes, of which 16,469 are protein-coding.

    A mean of 139,988 high-confidence pairs per tissue, ranging from 12,101 in bladder to 908,639 in whole blood. · Prioritization of high-confidence V2G…

  4. ShownMost high-confidence pairs are restricted to few tissues. 87% are classed as tissue-specific (not shared across 10 or more tissues), and sharing is higher among related tissues such as brain regions or blood-related tissues.

    13% of high-confidence pairs were shared across at least 10 tissues. In one example, rs7537004-CELA3B scored 10 in pancreas and had no high-confidence support in any other tissue. In another, rs2247303-RPL12 was high-confidence in all 50 tissues and scored 10 or more in 28 of them. · V2G evidence integration for high-confidence V2G…; Tissue-specific V2G associations

  5. SuggestedIn a worked example, mV2G ranks GLP1R as the top candidate gene for rs2268647, an intronic variant associated with body mass index. Support is strongest in pancreas, with further support in cardiac tissues.

    Several evidence classes, not a single dataset, support the pancreatic call. The text gives no per-tissue scores; these are shown only in a figure. The authors take this as genetic support for the clinical benefit of GLP-1 receptor agonists, which is not tested here. · Case study: mV2G prioritizes rs2268647-GLP1R…

Methods, data and reproducibility

Common variants (minor allele frequency >0.01, from FAVOR) were paired with all GENCODE v47 genes within 1 Mbp and annotated with 24 evidence types: xQTLs, enhancer-gene predictions, 3D chromatin contacts, CRISPR perturbations and three tissue-agnostic combined scores. A pair's score per tissue counts its supporting evidence types; results are served from an SQLite database via a PHP/JavaScript website.

  • Common variants from FAVOR annotations (IGVF Data Portal curated set), filtered to the FAVOR COMMON category, minor allele frequency >0.01 and alleles of 50 bp or less · human · n = 15,782,471 variants retained from 1,052,126,302 · IGVFDS5361GWAK
  • GENCODE Release 47 gene annotation · human · n = 74,628 genes (19,355 protein-coding, 55,273 non-protein-coding)
  • Candidate variant-gene pairs: all genes overlapping a 1 Mbp window around each variant · human · n = 425,576,568 pairs
  • Tissue-specific evidence types (xQTLs, enhancer-gene predictions, 3D chromatin interactions, CRISPR perturbation data) across 50 tissues used in population xQTL studies such as GTEx · human · n = 21 evidence types; 50 tissues
  • Tissue-agnostic variant-to-gene combining strategies applied to all tissues: cS2G, Open Targets, GeneHancer · human · n = 3 strategies
  • Genome browser tracks: ENCODE SCREEN candidate cis-regulatory elements and the Roadmap Epigenomics 15-state ChromHMM model for the matching tissue · human
Methods & tools
FAVOR, IGVF Data Portal, GENCODE v47, GTEx, cS2G, Open Targets, GeneHancer, ENCODE SCREEN, Roadmap Epigenomics ChromHMM, IGV, SQLite, PHP
Reproducibility gaps
No code or pipeline repository is linked. Processed data are available only through the website's query and download functions. Third-party inputs must be obtained from their original repositories.
Limitations (4)
  • Resolution is tissue-level only. The authors name extending the framework to individual cell types and cell states, using single-cell and spatial data, as an important future direction. · stated by the authors
  • Coverage depends on the datasets currently integrated. The authors plan future releases that add datasets, disease-specific regulatory evidence and experimentally validated pairs to increase the number and confidence of associations. · stated by the authors
  • The score is an unweighted count of evidence types, and tissue-agnostic evidence adds to every tissue. Data-rich tissues therefore collect more calls (whole blood 908,639 high-confidence pairs vs 12,101 in bladder), so scores partly track how much data exists and may double-count correlated sources. · noted in this summary
  • No benchmark against known causal genes, or against cS2G, Open Targets or GeneHancer, is reported. The cutoff rests on a single-gene-fraction proxy, and the only biological support is anecdotal (CELA3B, RPL12, GLP1R). · noted in this summary

Read this if you need tissue-resolved candidate target genes for noncoding GWAS variants and want a queryable, downloadable table of per-tissue evidence counts across 50 human tissues.

Citations
0

Abstract

The multiomic Variant-to-Gene (mV2G, https://mv2g.hbliulab.org) is a comprehensive atlas that integrates diverse functional genomic evidence to prioritize tissue-specific variant-to-gene (V2G) associations. While genome-wide association studies (GWAS) have identified millions of associations between genetic variants and diseases, translating these findings into biological mechanisms remains challenging because >90% of variants reside in noncoding regions. Existing V2G resources provide complementary regulatory evidence but are fragmented and often lack tissue-specific interpretation. To address this challenge, we constructed the mV2G atlas by integrating 24 types of functional genomic evidence across 50 human tissues, including molecular quantitative trait loci, enhancer-gene predictions, three-dimensional chromatin interactions, and experimental validation. The atlas contains 188,634,118 evidence-supported V2G pairs involving 13,618,039 variants and 69,521 genes. We further developed a unified tissue-specific V2G prioritization framework and prioritized 1,530,420 high-confidence functional V2G pairs involving 1,131,316 unique variants, with 87% exhibiting tissue-specificity. The mV2G atlas provides searchable variant- and gene-centered interfaces, an interactive browser for visualizing variants, target genes, cis-regulatory elements, and chromatin states, as well as downloadable datasets. By integrating complementary regulatory evidence into a unified framework, mV2G provides an accessible resource for interpreting the functional and phenotypic impact of genomic variation in relevant tissues for human diseases. GRAPHICAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=78 SRC="FIGDIR/small/743997v1_ufig1.gif" ALT="Figure 1"> View larger version (28K): org.highwire.dtl.DTLVardef@ac8319org.highwire.dtl.DTLVardef@1d31920org.highwire.dtl.DTLVardef@169225org.highwire.dtl.DTLVardef@1d4dc02_HPS_FORMAT_FIGEXP M_FIG C_FIG

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISSun '21Li '24Cheng '25Sooknah '26Rubin '21Fabiha '24
54% semantic matchGenetics
Quan Sun, Cheynna Crowley, Le Huang+9 more·Human Genetics and Genomics Advances·Oct 12, 2021· 15 citations