Structural insights on the honeybee immunity protein Vitellogenin, obtained with AlphaFold, are being applied to conservation of endangered bees and AI-assisted breeding programs.

Judged as of 2025-11-25. The claim has two parts: (a) AlphaFold-derived structural insights on honeybee Vitellogenin are being applied to conservation of endangered bee populations; (b) they are guiding the development of AI-assisted breeding programs for healthier pollinators. True if published research, project or organisation announcements (e.g. from the Vitellogenin research group or partner bee-health organisations) show both applications in progress at that date. Mostly true if both are under way but at an early or pilot stage, or if the conservation work concerns bee populations generally rather than specifically endangered ones. Mixed if only one of (a) or (b) is documented. Mostly false if the structural work exists but the evidence shows only plans or aspirations, with neither application actually under way. False if no such Vitellogenin structural work used AlphaFold, or if the evidence shows the applications do not exist.

Set before the research began.

1. E1 Project USDA Agricultural Research Service (Honey Bee Breeding, Genetics, and Physiology Research)

   “We targeted the protein Vitellogenin (Vg) that has several beneficial functions in lifespan, stress resistance and immunity as a test case. We successfully selected a breeding stock of honey bees homozygous for a reference Vg protein variant. This population was used in a breeding program to breed stocks with Vg protein variants with differences in lipid and zinc carrying abilities.”

2. E2 Project USDA Agricultural Research Service (Honey Bee Breeding, Genetics, and Physiology Research)

   “We progressed in pioneering the use of AI-assisted prediction of protein folding and function to develop a new pipeline for molecular-based selection of honey bees (4.4B).”

3. E3 Publication USDA Agricultural Research Service · 22 August 2025

   “In this article, we report that we were able to successfully breed for that standard protein product. This allowed us to speed up and decrease the cost of honey bee breeding.”

The USDA ARS material shows a Vitellogenin breeding program and a selection pipeline that uses AI-assisted prediction of protein folding [E1][E2][E3]. However, it does not name AlphaFold or link the work to AlphaFold-derived insights, and no item addresses conservation of endangered bee populations. Neither part of the claim is therefore documented as the criteria require.