Deep Origin's computational platform rationalizes molecular glue compound that potently kills DLBCL cells

Therna Biosciences appointed Kevin Green, former Chief Business Officer of Inceptive, as Chief Operating Officer and added two AI technical leaders to its advisory team, while the NIH Common Fund selected the company as a Phase I winner of its Targeting RNA in Disease with Novel Technologies Challenge. Green, who joined Therna in August and brings more than 20 years of experience across AI-enabled drug discovery and RNA therapeutics, leads the company's operations and business development as its RNA-Logix platform moves into partnered and internal programs. The new AI advisors are Dr. Dave Burke, the first Chief Technology Officer of Arc Institute and previously Vice President of Engineering for Android at Google, and Mr. Harsh Mehta, co-founder and CTO of Mirendil and previously a Senior Research Scientist at Anthropic and a researcher at Google. Therna is one of 25 Phase I winners of the three-phase TRDNT Challenge, which offers up to $13.1 million in total awards as teams develop and validate their technologies through 2027. Therna's winning proposal introduces a co-scientist agent built on RNA-Logix that designs and optimizes sequence-specific activating antisense oligonucleotides for ultra-rare haploinsufficiencies.

insitro announced that the advanced machine learning models it built with Eli Lilly and Company to predict in vitro and in vivo properties of small molecules are now available to certain biotech companies through Lilly TuneLab. The models were trained on multi-species preclinical data that Lilly generated over decades of research from hundreds of thousands of unique molecules, offering an alternative to the industry standard method of extrapolation from in vitro assays. The collaboration, announced in September 2025, paired insitro's computational and AI/ML expertise with Lilly's proprietary preclinical dataset of in vitro and in vivo measurements across compounds with established ADMET properties. The in vivo PK models are a critical component of insitro's TherML platform for small molecule discovery, alongside machine learning and physics-based in silico screening, affinity models powered by proprietary adaptive DNA-encoded libraries generating data up to 1B compound scale, and an active-learning medicinal chemistry design engine. Lilly TuneLab is part of Lilly Catalyze360, alongside Lilly Ventures, Lilly Gateway Labs, and Lilly ExploR&D, and is built on a federated learning infrastructure hosted by a third-party provider.

Profluent announced a collaboration with BioMarin Pharmaceutical Inc. to design and optimize novel enzyme therapies for BioMarin's pipeline. Under the agreement, Profluent will apply its foundational AI models to generate novel potential drug candidates, and BioMarin will screen the results with the goal of eventually advancing the most promising designs. Profluent will design and optimize enzyme candidates using its suite of proprietary protein language models, which draw on the Profluent Protein Atlas, described as the largest protein sequence dataset in the world with more than 115 billion unique sequences. BioMarin will use Profluent's AI models to rapidly explore protein designs and identify candidates with desired activity and stability profiles, then advance the most promising into preclinical testing. Pooja Agarwal, Vice President and Head of Metabolic Conditions Research at BioMarin, said the collaboration pairs Profluent's de novo protein design approach with BioMarin's decades of enzyme therapy expertise, while Profluent CEO Ali Madani said the partnership aims to accelerate BioMarin's drug discovery pipeline.

Ginkgo Datapoints, an offering of Ginkgo Bioworks, and Apheris GmbH announced that the Antibody Developability Consortium has launched with AbbVie, argenx, Lundbeck, and Takeda as founding members. The industry collaboration aims to build the field's largest standardized antibody developability dataset, with each founding member contributing proprietary antibody sequences and Ginkgo Datapoints filling any remaining capacity from publicly available sources to reach 10,000 antibodies in total. Using Apheris' federated infrastructure, members can train, benchmark, and refine AI models on the full consortium dataset without exposing their raw proprietary sequences to other members, and they retain ownership of the sequences and assay data they contribute. Ginkgo Datapoints leads the scientific design and execution, including sequence selection, antibody production, and high-throughput wet-lab characterization, and trains a foundation antibody developability model within Apheris' secure environment. Charlotte Deane of the University of Oxford and Peter Tessier of the University of Michigan will provide independent scientific oversight, and the consortium's initial dataset is targeted for delivery to members by early 2027.