The Gene Callability Atlas

A gene can read perfectly and still hide the answer.

An aggregation of sequencing blind-spot knowledge across 19,250 human genes, from 20 public sources, with evidence behind every flag.

Try:
◎
Gene not found

This symbol is not in the Blindspot dataset, which carries the 19,250 genes with a RefSeq Select transcript on chromosomes 1–22, X and Y.

Request full access
ⓘExon models from RefSeq Select; difficult-region overlap and intervals from the GIAB stratifications; category flags from the reference-assembly comparison; gene validity and dosage from ClinGen. Every figure is traceable to a documented source.
The problem
1 in 3human genes have at least one documented blind spot in their coding sequence, promoter or copy number
1 in 13human genes have coding sequence that standard short reads can’t reliably call
1 in 5pathogenic ClinVar variants sit in a region GIAB flags as difficult to sequence
10,657coding exons lie entirely in repeated sequence, where a short read may land on the wrong copy

Measured from the Blindspot database on 26 October 2026. How each figure is measured.

Early access

Ready to see the blind spots
in your panel?

Every gene opens with a free account: sign in with your email. For lab or team use, or to join the panel-report beta, leave your email and role and we will be in touch.

The evidence base →
Blindspot Genetics is an informational and educational reference tool. It surfaces published, gene- and assembly-level sequencing limitations to support test selection and clinical understanding. It does not provide medical advice, make patient-specific recommendations, or replace confirmatory testing, laboratory judgment, or genetic counseling. Clinical decisions remain the responsibility of the ordering provider and testing laboratory. Every figure is traceable to peer-reviewed literature or a documented internal pipeline.