Modern genomics is built on human population data
Variant interpretation today depends on human-cohort observations — allele frequencies, clinical classifications, and predictive scores trained on annotated variants. While these resources are crucial, they are fundamentally constrained by the limits of human diversity and sample size. A variant that is rare in humans may still be benign. A variant that has conflicting ClinVar entries may simply lack evidence. Prediction models provide estimates, not biological truth.
Human data tells us what we observed.
Evolution tells us what biology allowed.
Evolution ran the experiment long before we did
Across 80 million years, primates accumulated millions of natural substitutions. Some changes appeared independently across multiple lineages. Others were never observed in any species despite millions of opportunities. These patterns are not random — they are the outcome of constraint, tolerance, and functional necessity.
If a substitution recurred in multiple primates, biology demonstrated tolerance.
If a site remained unchanged across all 55 genera, biology showed constraint.
This deep-time record is the closest empirical truth set available for functional interpretation.
CodeXome transforms this record into a searchable, quantitative framework that researchers can use in seconds.
Why evolution is the missing layer
Natural variation reveals tolerated substitutions
When a missense change occurs independently across primate lineages, the conclusion is straightforward:
that substitution does not disrupt function.
Evolution has already validated its viability.
Constraint highlights functional sensitivity
Positions that never change across 55 primate genomes — despite millions of years — signal functional necessity. These sites behave like “biological red lines,” where substitution is highly unlikely to be benign.
Deep Time Ancestry creates a measurable index of tolerance
CodeXome’s Deep Time Ancestry Score quantifies recurrence and absence into a numerical indicator that supplements population frequency and prediction scores.
This is empirical evidence, not model output.
What primate recurrence looks like in real data
CodeXome aligns human coding sequences to 55 primate genomes, revealing:
- White squares: identical to human
- Colored squares: evolutionary substitutions
- Recurring substitutions: evidence of tolerance
- Unchanged sites across all species: evolutionary constraint
These patterns illuminate functional regions invisible to human-only datasets — including IDRs, motifs, structural domains, transmembrane loops, enzymatic pockets, and binding regions.
Questions evolution can answer that modern tools cannot
- Is this variant tolerated by biology?
- Does it fall in a deeply conserved region?
- Does recurrence occur within the same chemical class?
- Is the variant human-unique or multi-lineage?
- Has selection ever allowed this substitution across millions of years?
Population datasets quantify what’s present.
Evolution reveals what was possible.
When both are combined, researchers obtain a more complete picture of functional significance.
How CodeXome complements today’s genomic tools
Current Genomic Tools
Basis: Human population frequencies
Scope: Human-only variation
Output: Prediction scores
Bias: Limited by human ancestry representation
Evidence Source: Observed variation
Interpretation: Predictive
CodeXome
Basis: Evolutionary recurrence across primates
Scope: 55 lineages over 80 million years
Output: Deep Time Ancestry Score (DTAS)
Bias: Balanced across species
Evidence Source: Empirical tolerance or constraint
Interpretation: Evolution-tested
Examples from real genes
CFTR
Over 99% of CFTR pathogenic sites appear nowhere in primates, reinforcing their functional importance.
SCN5A
Evolutionary recurrence resolves 79% of SCN5A VUS as likely benign — providing clarity where human datasets offer uncertainty.
PKD1
Evolutionary inference aligns with ClinGen expert consensus, particularly in difficult-to-classify missense regions.
These are not statistical estimates — they are biological observations.
Evolution agrees with functional truth
Across disease areas — neurology, cancer, immunology, metabolic disorders — primate recurrence and conservation map directly onto known functional domains and annotated pathogenic sites. The result is a biological framework that supports, clarifies, or corrects population-based predictions.
Evolution is not a model.
Evolution is evidence.
See evolution in action on your gene
A CodeXome trial includes:
- Full access to the Deep Time Ancestry Database
- Gene Profile exploration
- Variant Analysis with shared/unique filtering
- Evolutionary tolerance scoring
- Integrated overlays with ClinVar, UniProt, gnomAD
Every gene has an evolutionary story. Before CodeXome, most were hidden.
