See how pathogenic your mutations are according to AlphaMissense based on your 23andme raw data

Snakemake pipeline for visualizing AlphaMissense pathogenicity score by UniProtID. Analysis of Asparagine Synthetase predictions.

Elementary interfaces to AlphaMissense resources, using BiocFileCache, Rsamtools, GenomicRanges

The Amissense Tool analyzes and visualizes AlphaMissense pathogenicity scores, integrating AlphaFold structures and ClinVar data. It offers automated pipelines, visualizations, and versatile command-line utilities.

Calibrate functional/in-silico variant scores into ACMG clinical evidence strengths (ClinGen-SVI), with LDLR/AlphaMissense worked example

Five Google ADK / Agent Builder agents watching genomic evidence (ClinVar, gnomAD, AlphaMissense) synced via the Fivetran MCP and, the moment a Variant of Uncertain Significance is reclassified, recompute a calibrated ACMG posterior to draft the patient recontact + family cascade no system sends today. Draft-only, FHIR R4, human-in-the-loop.

Alpha Missense Scores provides precomputed JSON files with AlphaMissense pathogenicity scores for protein variants. Derived from DeepMind's TSV dataset, these JSON files are can be accessed via HTTP queries. This repository allows efficient data retrieval, bypassing the need for local processing of the original TSV files.

Machine-learning triage of LDLR variants of uncertain significance using predictor concordance and ACMG-aligned evidence mapping

Annotate any PDB structure with AlphaMissense pathogenicity scores at the residue level.

Validation of AlphaMissense

Evaluating AlphaMissense pathogenicity scores for PDAC diagnostics and molecular subtyping, using TCGA-PAAD and 1000 Genomes data

Turn raw 23andMe/AncestryDNA data into a private, local HTML health report — clinical variants (ClinVar, AlphaMissense), pharmacogenomics (PharmGKB), and polygenic risk scores (PGS Catalog), plus an optional AI summary. Runs 100% locally; your genome is never uploaded.

Elementary interfaces to AlphaMissense resources, using BiocFileCache, Rsamtools, GenomicRanges

Supplementary Data for Pillai et al., 2026.

Computational pipeline for prioritizing structurally tractable pharmacologic rescue candidates across KCNQ1-KCNQ5.

Machine-learning triage of LDLR variants of uncertain significance using predictor concordance and ACMG-aligned evidence mapping

CFTR variant pathogenicity annotation using CFTR2 and AlphaMissense. AUC 0.946 on 292 labelled variants. 7 unclassified variants predicted pathogenic with population frequency support.