Expanding the utility of variant effect predictions with phenotype-specific models - Nature Communications V2P predicts variant pathogenicity conditioned on disease phenotypes across top-level Human Phenotype Ontology categories. This approach shows promise for phenotype-specific estimation of variant effe...

Check out our new paper introducing V2P — a method that predicts both variant pathogenicity and disease phenotype across 23 HPO categories. With @itanlab.bsky.social, David Stein, and many other great collaborators
www.nature.com/articles/s41...
www.v2p.ai

Lastly, we dissect the biology of pathogenic variants by disease domain using feature selection: discriminative annotations for each top-level HPO group, revealing which signals are shared across domains and which are phenotype-specific.

Additionally, we show in real patients' exomes that the known causal variant has a median rank of #2 when prioritizing with the V2P phenotype-specific score(s) matching the patient’s phenotype.
@casanovalab.bsky.social

V2P webserver: v2p.ai

V2P (variant-to-phenotype) is live: nature.com/articles/s41...
To our knowledge, first genomewide SNVs+indels model jointly predicting pathogenicity + disease domain (23 HPO groups; e.g. cardiac/immune/metabolic).
Great work by David Stein in collaboration with @schlessingerlab.bsky.social

New AI tool identifies not just genetic mutations, but the diseases they may cause Scientists at the Icahn School of Medicine at Mount Sinai have developed a novel artificial intelligence tool that not only identifies disease-causing genetic mutations but also predicts the type of disease those mutations may trigger.

A new AI tool links genetic mutations to specific disease types, enhancing the speed and accuracy of genetic diagnostics and supporting the discovery of targeted treatments for complex conditions. doi.org/hbfn92

itan-lab / Digenic Analysis · GitLab GitLab.com

In this project, Ece developed a robust digenic case-control association framework, which we highly recommend applying to various disease groups to uncover missing heritability. Guideline for the process: gitlab.com/itan-lab/dig...

Our new publication on the digenic architecture (two causative genes in a single patient) of congenital heart disease is now online: www.sciencedirect.com/science/arti...
Congrats to Ece Kars who led this work, and thanks to Bruce Gelb & the PCGC consortium for the collaboration.