alphagenome-variant-impact-score

Score and annotate genetic variants with AlphaGenome Variant Impact scores via the Atlas API.

Updated Sep 9, 2026
One-click install
npx skills add https://github.com/xlinh2301/EditCTC --skill alphagenome-variant-impact-score-xlinh2301
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: alphagenome-variant-impact-score
Source: https://github.com/xlinh2301/EditCTC/tree/main/.agents/skills/alphagenome-variant-impact-score
Command: npx skills add https://github.com/xlinh2301/EditCTC --skill alphagenome-variant-impact-score-xlinh2301

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires absl-py, alphagenome, anndata, numpy, pandas, polars, pyarrow, pysam, python-dotenv, tabulate, and includes scripts (resource) components.

What problem does it solve? Prioritizing which genetic variants are functionally important from large callsets is slow and requires integrating many genomic data modalities. This Skill scores, annotates, and ranks variants using AlphaGenome Variant Impact (AVI) models through the AlphaGenome Atlas API, returning calibrated Phred scores and biological feature attributions. ## Core Features & Use Cases - Variant Querying & Scoring: Query single or multiple variants in chr:pos:ref>alt format to get AVI Phred scores, quantiles, and 18 biological feature attribution weights. - VCF/Tabular Annotation: Annotate VCF, CSV, TSV, or Parquet callsets in Ensembl VEP CSQ format and rank top candidate variants. - Saturation Mutagenesis Scans: Score all possible single-nucleotide substitutions across a genomic window to find functional hotspots. - GENCODE v46 Gene Annotations: Extract exon, CDS, UTR, and splice junction coordinates for genes and transcripts. - Use Case: Given a VCF of 1,000 rare variants from a sequencing study, annotate each with AVI scores, filter to Phred >= 20 (top 1% genome-wide), and identify the top splicing-disrupting candidates with Atlas deep-links. ## Quick Start Ask the assistant to score the variant chr9:128225994:G>A with the AlphaGenome AVI skill and show its Phred score and top contributing biological modality.

Frequently Asked Questions about alphagenome-variant-impact-score

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I score a genetic variant with AlphaGenome Variant Impact?▼

Run the query subcommand with a variant in 1-based chr:pos:ref>alt format, for example: uv run scripts/alphagenome_atlas_avi.py query "chr9:128225994:G>A" --format json. The output includes the AVI Phred score, quantile, and top contributing biological modality.

How do I annotate a VCF file with variant impact scores?▼

Use the annotate subcommand with --input and --output flags to stream a VCF through the Atlas API. Annotations are injected in Ensembl VEP CSQ format with AVI_PHRED, AVI_RAW, AVI_QUANTILE, AVI_TOP_PERCENTILE, and AVI_TOP_FEATURE fields, and throughput is about 10 variants per second.

What does the AlphaGenome AVI Phred score mean?▼

The Phred score is a calibrated impact measure computed as -10 * log10(1 - quantile). Phred >= 40 indicates the top 0.01% of genome-wide SNVs, Phred >= 20 the top 1%, and Phred < 10 the bottom 90%.

Can AlphaGenome AVI scores be used for clinical diagnosis?▼

No. AlphaGenome AVI is strictly a research tool, and its terms of service prohibit clinical use. Scores reflect predicted molecular and functional impact such as splicing disruption or transcription factor binding changes, never clinical diagnoses or treatment recommendations.

Why does the gtf command take a long time on first run?▼

The gtf subcommand ingests a 318 MB GENCODE v46 feather dataset requiring about 4.37 GB of RAM. Cold-start downloads take 5-10 seconds on fast networks but up to 60-90 seconds externally, while cached queries run in 4-6 seconds.

What input formats does variant annotation support?▼

The annotate subcommand accepts VCF and VCF.GZ files parsed with pysam, plus CSV, TSV, and Parquet tabular files containing chromosome, position, ref, and alt columns. Outputs can be written as VCF, TSV, CSV, Parquet, or JSON.