Here's something a bit different that we've been noodling around with! Nearly 10y ago we established a panel of cell lines from Papuan donors, which let us check predictions from AlphaGenome against experimental data. If you've followed the field you'll know that expression prediction is very tough
Using sequence-to-function models to interpret archaic hominin introgression https://www.biorxiv.org/content/10.64898/2026.08.31.748430v1