insitro and Bristol Myers Squibb Add New AI-Found Targets in ALS Research

The ongoing collaboration between insitro and Bristol Myers Squibb has been extended to include additional AI-discovered therapeutic targets for amyotrophic lateral sclerosis (ALS), with early experimental validation in human neuron models. The nomination of the two new targets triggered a $10 million milestone payment to insitro.

Two newly nominated targets, ALS-2 and ALS-3, were identified using insitro’s Virtual Human (ISH) platform, which integrates large-scale human cell data with machine learning to map disease biology. The additions build on an earlier target, ALS-1, nominated by Bristol Myers Squibb in late 2024.

insitro, founded by Daphne Koller, is building an end-to-end drug discovery platform that combines human cell models with machine learning to identify and develop new therapies. Its ISH and POSH systems generate large-scale phenotypical and multi-omics data using iPSC-derived disease models, CRISPR perturbations, and high-content imaging, which are then used to train cell-level models of disease. 

In drug design, insitro’s ChemML and TherML platforms integrate experimental and computational data to design small molecules, oligonucleotides, and biologics. Alongside Bristol Myers Squibb, the platform supports both internal programs and partnerships with big pharma companies such as Gilead and Eli Lilly.

Under the agreement, both companies are pursuing multiple therapeutic strategies against ALS-1. insitro is advancing an oligonucleotide-based program, while Bristol Myers Squibb is developing a small molecule approach against the same target. This parallel modality strategy is aimed to increase the likelihood of clinical translation by addressing the same biological mechanism through different intervention types. 

In the US, the main approved ALS drugs mostly slow decline or address a narrow genetic subgroup. That leaves a field with few durable successes and very high demand for new mechanisms, especially ones that could apply beyond rare genetic subtypes. FDA’s 2023 approval of tofersen, a genetically precise antisense therapy limited to the SOD1-mutated subgroup, showed that RNA-targeted, genetically precise approaches can reach patients in ALS, even if only for a small subset so far.

The BMS-insitro discovery work focuses on TDP-43 mislocalization, a pathology reported in about 97% of ALS cases, which is why so many groups now treat it as the closest thing the field has to a unifying disease mechanism—-and a possible route to developing a drug that could be relevant to a much broader group of patients.

insitro’s platform is used to find the biological processes that lead to this problem by studying disease effects in human-derived cell models. In lab-grown motor neurons, adjusting these targets helped restore normal neuron structure and improved RNA processing, including reducing abnormal RNA fragments and bringing back full-length transcripts—signals that cells are functioning more “normally”.

Topic: Biotech Ventures