Collaborations Pharmaceuticals highlights a decade of AI case studies in drug research

Collaborations Pharmaceuticals says a decade of AI work has produced a set of case studies showing how its tools have been used for drug discovery, toxicology, polymers and consumer product research. The company says the examples demonstrate safe, real-world AI applications for human health and other fields.

Why it matters: - Collaborations Pharmaceuticals is framing AI as a practical research tool for pharmaceutical, chemical and consumer product work, not just as foundation-model technology. - The company says the case studies show AI can support safer decision-making in drug discovery, toxicology and product development. - The examples are meant to show real-world applications that could affect healthcare outcomes and other human health uses.

What happened: - Collaborations Pharmaceuticals, Inc. said it has summarized a decade of AI work into a dozen brief case studies. - The Raleigh-based company said the projects reflect more than 10 years of using AI in research and development. - CEO Sean Ekins said the company has worked with outside clients across the pharmaceutical, chemical and consumer product industries. - Ekins said many of the projects addressed problems those companies could not handle with their own internal capabilities.

The details: - Collaborations Pharmaceuticals said it developed AI software products including Assay Central, MegaTox, MegaTrans, MegaSyn and MegaAChE. - The company said NIH grants and matching funds from the One NC small business program helped fund those tools. - Collaborations Pharmaceuticals said it validated the software using internal chemistry and biology laboratories and external collaborations. - The company said its client work has included toxicology predictions, off-target prediction, designed PROTACs, polymers, PFAS alternatives and PK/PD models. - The company also said it has built bespoke tools for clinical trial data modeling, GC/MS spectra prediction and clinical trial success prediction. - The case studies listed by the company include de-risking a large molecule through predictive ADME/Tox and metabolism profiling, custom multi-target models and patent screening, and generative design of selective ADORA3 agonists for an anticancer target. - Additional case studies cover off-target effects of drug candidates, repurposing approved drugs as specific cytokine inhibitors, identification of dual kinase inhibitors and determining the Kd of a molecule against PPT1. - The list also includes designing new polymers for human health applications, assessing endocrine disruption potential of consumer product ingredients, designing new prodrugs with predicted ADME/PK properties, improving hERG and solubility for PROTACs, predicting off-target liabilities for a PROTAC molecule and developing an algorithm to predict ocular bioavailability.

Between the lines: - The company is using the case studies to counter skepticism around AI by pointing to lower-profile, specialized applications with direct research value. - The emphasis on confidentiality suggests most of the work has been fee-for-service and not publicly visible. - The examples suggest small AI teams can be competitive when they focus on narrow scientific problems rather than broad foundation models.

What's next: - Collaborations Pharmaceuticals said it expects to keep taking on bespoke client projects. - Ekins said the company wants to continue sharing case studies from future work. - The company is positioning itself for more challenging projects in drug discovery, safety assessment and related scientific modeling.