Sun · American journal of human genetics 2023 · cross-sectional bibliometric audit · n=430 polygenic scores sampled (GWAS catalog count not stated)

eXclusionarY: 10 years later, where are the sex chromosomes in GWASs?

Level 4 - case-series / case-control

Cross-sectional meta-research audit of public genetic databases (by design analogy, non-clinical).

PubMed 37267899 · doi:10.1016/j.ajhg.2023.04.009 · record verified 2026-08-26

What was done

Authors re-surveyed the genetic research landscape to evaluate whether recommendations to include sex chromosomes in genome-wide association studies (GWASs) were adopted. They audited genome-wide summary statistics published in 2021 in the NHGRI-EBI GWAS Catalog for X and Y chromosome coverage, calculated the physical-length-normalized density of significant findings published through November 2022, compared decadal study growth rates between the X chromosome and autosomes, and evaluated a sample of 430 scores from the PolyGenic Score Catalog.

What was found

Among 2021 NHGRI-EBI GWAS Catalog summary statistics, 25% included X-chromosome results and 3% included Y-chromosome results. Normalizing by physical length, the X chromosome had ~1 study/Mb with genome-wide-significant findings, compared to ~6 to ~16 studies/Mb for chromosomes 4 and 19. Over the preceding decade, X-chromosome studies grew at ~0.012 studies/Mb/year versus ~0.086 studies/Mb/year for autosomes. Zero percent (0/430) of sampled polygenic scores contained weights for sex chromosomal SNPs. Studies reporting X-chromosome findings displayed extreme analytical and reporting heterogeneity.

Why it matters

Despite established guidelines, sex chromosomes remain widely omitted from genomic studies and risk scores, rendering most modern GWASs effectively autosome-wide rather than genome-wide.

Limits

The total number of GWAS summary statistic files audited from 2021 is not specified in the abstract. Data were derived exclusively from public repositories (NHGRI-EBI and PolyGenic Score catalogs), which may reflect repository submission practices rather than all published genomic studies. Methodological details of the noted analytical heterogeneity were not quantified in the abstract.

Cited by