Every amino acid substitution that survived 80 million years of primate evolution was tested by natural selection in organisms biochemically near-identical to us. If it persisted across 55 genera, biology already approved it. If it never appeared, biology hasn't permitted it.

CodeXome captures that record. Not as a model's prediction, but as a direct empirical observation, and delivers it as residue-level annotation that drops into existing genomic interpretation pipelines.

The result is a genuinely new evidence axis. Orthogonal to population frequency databases. Orthogonal to ML predictor scores. Orthogonal to deep vertebrate conservation tools. Adding CodeXome to a pipeline adds something none of those categories can provide: a biological fact about what nature actually tested and tolerated, at the specific amino acid position in question.

What this enables:
— 50–75% reduction in candidate variant lists before any manual review begins
— 10–40% of Variants of Uncertain Significance identified as likely benign by evolutionary evidence, across ~14,000 ClinVar genes
— Population-agnostic evidence that carries identical evidentiary weight regardless of patient ancestry

Validated at scale:
100% concordance with expert panel classifications across the ClinGen 52-gene study (2,689 SNVs) and ENIGMA/BRCA Exchange (10,699 BRCA1/2 variants).

Who we work with:
Research genomics labs, rare disease programs, biotech target validation teams, and bioinformatics cores integrating new annotation sources into existing pipelines.

CodeXome is a product of Cornerstone Genomics. Research use. Residue resolution. 19,244 genes. GRCh38.

Cornerstone Genomics

Org chart

Teams

Offices