Roshni Patel

roshnipatel.bsky.social

06 October 2026 21:56

Would love to hear any feedback, and for more reading on the impact of the 2023 NASEM report, check out our preprint surveying early-career researchers, led by @jayatirsharma.bsky.social and @christacaggiano.bsky.social, from earlier this year: www.biorxiv.org/content/10.6...

Roshni Patel

roshnipatel.bsky.social

06 October 2026 18:06

But, this also illustrates the importance of bottom-up change: decades of work by individual researchers across the field have systematically improved our use of population descriptors in human genetics and genomics. We need to continue to invest in these efforts -- ethics and rigor are intertwined!

Roshni Patel

roshnipatel.bsky.social

06 October 2026 18:06

To be clear, this is good news! Over the last 20 years, human genetics GWAS have consistently followed all 2023 NASEM guidelines more thoroughly. (One notable exception is community engagement, which remains poor.)

Roshni Patel

roshnipatel.bsky.social

06 October 2026 18:06

Surprisingly, we also find that improvements in the use of population descriptors far predate the publication of the NASEM report -- suggesting these changes are not the result of any single intervention but rather the cumulative effects of tiny shifts in research practices and community norms

Roshni Patel

roshnipatel.bsky.social

06 October 2026 18:06

Here, we use the recommendations of the NASEM report as a rubric to evaluate the use of pop descriptors. We find that GWAS on datasets from outside the US or UK, or on non-European populations, tend to score better, suggesting these GWAS systematically use pop descriptors more accurately + ethically