Genomic Coordinates

What it does

Targets the quietest class of bug in genomics — coordinate errors that parse cleanly and return plausible but wrong answers.

• Format conversion: moves intervals between 0-based half-open formats (BED, bedGraph, PSL, genePred, bigWig) and 1-based inclusive ones (GFF3, GTF, VCF, WIG, Picard interval_list) using an explicit table. convert_coords.py also parses UCSC-style region strings and flags zero-length features as unrepresentable instead of silently mangling them.
• Variant normalisation: normalize_variant.py trims to parsimony and left-aligns indels against a reference FASTA, checks every REF against the FASTA, splits multi-allelic records, and compares two representations for equivalence.
• Assembly checks: check_contigs.py reads .fai, .chrom.sizes, VCF/SAM headers, FASTA, BED and GTF, identifies GRCh37 / hg19 / GRCh38 / T2T from chromosome lengths, and reports chr-prefix mismatches, length conflicts and missing contigs.
• File audits: audit_intervals.py finds the fingerprints of coordinate mistakes and exits 1 on fatal findings, so it doubles as a CI gate.
• Transcript/CDS/protein positions: explains HGVS numbering, the no-c.0 rule, GFF phase, and why conversion needs a transcript model rather than arithmetic.

Who it's for

• Bioinformaticians and pipeline engineers handling VCF, BED and GTF daily
• Anyone whose ClinVar/gnomAD joins return suspiciously few matches
• Teams untangling mixed GRCh37/GRCh38 pipelines
• Trainees who want a trustworthy reference for coordinate conventions

Examples

1. Off-by-one hunt: "My BED × GTF intersect looks shifted by 1bp" — the skill checks both conventions and audits the BED for 1-based rows.
2. Missing variant match: normalise chr1:7:CAC:C to chr1:2:GCA:G and confirm both describe the same deletion before joining to a database.
3. Build mismatch triage: run check_contigs.py --identify to distinguish GRCh37 from hg19 by mitochondrial length (16,569 vs 16,571 bp) before trusting any join.

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