Drug-discovery partnerships, safety pilots and research priorities

1. Earendil Labs Announces Research Collaboration with Genentech to Discover and Develop Therapeutic Bispecific Antibodies in Oncology¹

Earendil Labs announced a collaboration with Genentech to discover and develop bispecific cancer antibodies using its AI-driven biology platform. Earendil would lead discovery through early clinical development, with Genentech responsible for later global development and commercialisation. The announced terms include $55 million upfront and a total potential value above $1.5 billion, including contingent milestones, plus possible royalties. The transaction remains subject to closing conditions. This is a substantial commitment to model-guided biologics development; the headline potential value is not money already received or evidence that resulting treatments will succeed.¹

2. Drug Developers Deploy In Silico Drug Safety Prediction Models From Deep Origin’s ARPA-H-Funded PREDICTS Consortium²

Deep Origin says four organisations are implementing its PREDICTS consortium’s drug-safety models in a pilot programme. ImmVue, Synko, Sanford Burnham Prebys and SyzOnc will use the predictors to prioritise compounds, with laboratory results returned to assess performance and expand training data. The company reports that 62 of 77 planned ADMET models are available to participants. The operational step matters because predictions will encounter new experimental feedback. It does not establish improved clinical safety, and the separate virtual-human organ models described in the announcement remain in development rather than a completed validation result.²

3. Genialis and Inventia Life Science Collaborate to Link Patient Biology with 3D Pancreatic Cancer Models³

Genialis and Inventia Life Science announced a collaboration connecting AI-derived patient tumour profiles with controllable three-dimensional pancreatic cancer models. Genialis’ transcriptomic model will characterise the biological states represented by Inventia’s RASTRUM tumour–fibroblast co-cultures, helping researchers choose laboratory systems for testing therapeutic hypotheses. The companies describe linking profiles from the PanCAN SPARK cohort to these experimental models. The news is the partnership and research workflow, rather than proof that it predicts treatment benefit. Their announcement scheduled initial findings for the PanCAN summit; it does not by itself verify the presentation or clinical validation.³

4. Genesis Mission Frameworks: Advancing the Frontiers of American Discovery⁴

The U.S. Department of Energy announced publication of an advisory report on applying AI to major scientific challenges. Its three highlighted areas are predictive biology, fusion energy and domestic permanent-magnet capability. For chemistry and materials researchers, the proposals emphasise connected measurement, modelling and manufacturing, including computational searches for magnet compositions and systems for recovering critical materials. These are strategic recommendations within the Genesis Mission, not demonstrated breakthroughs or newly awarded grants. The report’s release clarifies proposed research directions, while its ambitious therapeutic and manufacturing goals still require execution and evidence.⁴

5. FSU chemist earns Florida Cancer Innovation Fund grant to create new pediatric cancer treatments from existing medicines⁵

Florida State University announced that Qing-Xiang Amy Sang received more than $500,000 from the Florida Cancer Innovation Fund for AI-assisted drug repurposing in brain cancer. The project will screen existing drug and disease information to prioritise candidates, then test them in laboratory systems including patient-derived tumour organoids. Rare childhood brain cancers are a central focus. The funding links computational selection to planned experimental follow-through in an underserved disease area. The university’s announcement describes a research programme: it does not report a newly validated treatment or establish that an existing drug benefits these patients.⁵