Editing Openai/69177fbb-14c4-800f-bc57-ccb5a1d88c95 (section)

=== - Polygenic score differences and cross-population patterns. Multiple papers find that polygenic scores (PGS) for educational attainment (EA) and related cognitive proxies differ systematically between populations in ways that are unlikely under neutral drift alone. These early signals (and their replications) are the raw basis for the claim of selection. Preprints<ref>{{cite web|title=Preprints|url=https://www.preprints.org/manuscript/201611.0047/v1/download|publisher=preprints.org|access-date=2025-11-17}}</ref> ===
* Ancient-DNA time transects. Ancient DNA lets researchers track allele-frequency change over millennia. Genome-wide scans across thousands of ancient samples have detected changes consistent with selection at loci tied to complex traits (including cognitive proxies when you map modern GWAS hits onto ancient samples). Mathieson et al. (2015) and follow-ups established the utility of this approach. PMC<ref>{{cite web|title=PMC|url=https://pmc.ncbi.nlm.nih.gov/articles/PMC4918750/|publisher=pmc.ncbi.nlm.nih.gov|access-date=2025-11-17}}</ref>
* Increasing GWAS power reduces noise. More recent, very large GWAS (e.g., Okbay et al. 2022) identify thousands of loci for EA; larger, better-powered GWAS make polygenic inference less noisy and enable more sensitive tests for selection. PubMed<ref>{{cite web|title=PubMed|url=https://pubmed.ncbi.nlm.nih.gov/35361970/|publisher=pubmed.ncbi.nlm.nih.gov|access-date=2025-11-17}}</ref>