AI Skill MartBrowse skillsPromptsProgramsLearnWhat is a Skill?
 
← Back to list

AlphaGenome Variant Impact

Look up, score, and mechanistically interpret regulatory and splicing effects of SNVs using DeepMind's AlphaGenome model and the AlphaGenome Atlas.

Data & AnalyticsAdvanced★ 55,210⑂ 4,861AI score 8/10Last updated: Oct 26, 2026

What it does

  • Atlas lookups: fetch precomputed AVI scores (raw value, genome-wide quantile, Phred) plus all 18 SHAP feature attributions for any GRCh38 single-nucleotide variant, by variant, VCF, or ~1 kb window.
  • Track-level mechanism: pull per-scorer scores for RNA-seq, DNase, ATAC, ChIP-TF, ChIP-histone, CAGE, PRO-cap, splicing, polyadenylation and contact maps, filtered by tissue ontology CURIE.
  • On-demand model calls: for indels, mouse (mm10), custom backgrounds or windows the Atlas doesn't hold, score directly with AlphaGenome, predict REF-vs-ALT tracks, and run in-silico mutagenesis.
  • Reporting discipline: always report raw score and quantile/Phred with scorer, track, biosample and gene; rank rather than threshold; treat "no appreciable change" as a valid answer.
  • Portal deep links: build Atlas website links per variant, locus, or gene offline, no key needed.

Who it's for

  • Genomics researchers prioritising non-coding candidates from GWAS fine-mapping or credible sets.
  • Analysts who must explain promoter, enhancer, splice-site, or chromatin-accessibility mechanisms.
  • Anyone handling VCFs who wants strict coordinate hygiene (1-based vs 0-based, GRCh37 vs GRCh38, rsID resolution).

Example uses

  1. atlas_query.py avi --input candidates.vcf --min-phred 20 -o avi.tsv to keep only top-1% variants.
  2. atlas_query.py scores --variant "chr22:36201698:A>C" --scorers RNA_SEQ DNASE --ontology UBERON:0001157 to resolve a colon-specific mechanism.
  3. Score indels or mouse variants with score_variants.py --organism mouse ..., then attach an atlas_link.py portal URL to every reported variant.

Research and theoretical modelling only; not a clinical or diagnostic tool.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/K-Dense-AI/scientific-agent-skills/HEAD/skills/alphagenome/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on AI Skill Mart.
Copy the skills/alphagenome folder from the GitHub repo K-Dense-AI/scientific-agent-skills into my ~/.claude/skills/alphagenome/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/K-Dense-AI/scientific-agent-skills.git /tmp/sas && mkdir -p ~/.claude/skills && cp -r /tmp/sas/skills/alphagenome ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Confirm Python 3.10+ (3.12/3.13 tested) and uv or pip are available.
  2. Install the SDK: uv pip install "alphagenome>=0.9.0" (includes the Atlas client).
  3. Get a free non-commercial API key at https://deepmind.google.com/science/alphagenome
  4. Add export ALPHAGENOME_API_KEY="your_key" to ~/.zshrc or ~/.bashrc and open a new shell. Never pass the key on the command line or commit it.
  5. Install the skill: 
    git clone https://github.com/K-Dense-AI/scientific-agent-skills.git /tmp/sas
mkdir -p ~/.claude/skills && cp -r /tmp/sas/skills/alphagenome ~/.claude/skills/
  6. Smoke test: cd ~/.claude/skills/alphagenome/scripts && python atlas_query.py scorers — this verifies both the key and network access to gdmscience.googleapis.com:443.
  7. Restart Claude Code and ask something like "prioritise the non-coding variants in this VCF with AlphaGenome and link each to the Atlas portal."
View source on GitHub ↗License: MIT