BOSTON– Ginkgo Bioworks (NYSE: DNA) announced a slate of new initiatives through its Datapoints platform to accelerate the use of artificial intelligence in biologics drug discovery. The company has partnered with Apheris to launch the Antibody Developability Consortium and separately introduced the AbDev AI Competition, both aimed at improving data infrastructure and advancing predictive modeling in antibody research.
The Antibody Developability Consortium will tackle one of the field’s central challenges: predicting and optimizing antibody properties early in R&D to boost downstream clinical and commercial success. Ginkgo will leverage its AI and high-throughput lab capabilities to generate purpose-built datasets, while Apheris will provide federated computing infrastructure that allows participants to collaborate securely without relinquishing ownership of sensitive data. The consortium is now enrolling member companies, with initial datasets and models for multiple antibody formats expected in 2026.
“The future of AI in drug discovery depends on creating environments where companies can collaborate without compromising their most valuable data,” said Robin Röhm, CEO and cofounder of Apheris. “With this new consortium, we’re bringing federated learning directly into antibody R&D, making critical datasets usable in ways that were never possible before.”
In parallel, Ginkgo is launching the first-ever AbDev AI Competition on the Hugging Face platform to establish benchmarks for antibody developability modeling. Running through early November, the competition will test model performance against standardized datasets, with winners set to receive up to $60,000 in prize value.
“One of the biggest barriers in antibody AI has been the lack of large, high-quality datasets on developability,” said Peter Tessier, professor at the University of Michigan and advisor to Ginkgo Datapoints. “Consortia and competitions like these are a crucial step toward closing that gap.”