Multi-scale haplotype signatures distinguish among evolutionary processes

SHADES, a deep learning framework for multi-scale haplotype data, distinguishes neutrality, selective sweeps, balancing selection, and adaptive introgression, outperforming contemporary methods on accuracy and power.

• Why it matters: Distinguishing among multiple evolutionary processes from genomic variation is still difficult, limiting reliable interpretation of adaptive signatures in population-scale data.
• What they did: SHADES integrates sequence differences and their frequencies across spatial resolutions to learn haplotype signatures, then is benchmarked against several existing methods for classification accuracy and power.
• The result: It stays robust to confounders with only modest performance changes, and on European human haplotypes it recovers known signals and identifies novel genes tied to immune defense, inflammatory regulation, and epithelial barrier biology.

AI summary of the abstract · may contain errors · not medical advice · read the paper.