Boehringer Ingelheim is committing more than $1 billion to a partnership with the AI biotech Envisagenics, betting that a machine-learning tool can unlock a hidden layer of cancer biology that has been largely overlooked as a source of drug targets: RNA splicing.
What RNA splicing is — and why it matters
When a gene is switched on, its DNA is first copied into RNA, which is then edited before it becomes a protein. That editing — splicing — can stitch the same gene together in different ways, producing multiple proteins from one gene. Cancer cells frequently hijack this process, generating abnormal “spliced” proteins that healthy cells don’t make. Those aberrant proteins are, in principle, ideal drug targets — they’re on cancer cells but not normal ones — yet they’ve been hard to find systematically. Boehringer’s oncology research head, Mark Petronczki, called alternative RNA splicing “a largely unexplored target space.”
What Envisagenics brings
Envisagenics built an AI platform called SpliceCore that screens more than 14 million alternative splicing events to identify which ones produce viable, tumor-selective targets. In effect, it uses computation to sift an enormous, noisy dataset for the rare splicing quirks that could be turned into therapies — a task impractical to do by hand. It’s a good example of AI’s most credible near-term role in drug discovery: not designing drugs from scratch, but finding the right targets in overwhelming biological data.
The deal — and what they’ll build
The agreement is a research collaboration and option deal valued at over $1 billion, combining upfront payments, research funding, option fees, milestone payments tied to development and approval, and royalties on future sales. The partners aim to develop precision therapies for solid tumors designed to selectively destroy cancer cells while sparing healthy tissue. The therapeutic formats on the table are the modern toolkit of targeted oncology: antibody-drug conjugates, multi-specific antibodies and T-cell engagers — all of which need a good target to aim at, which is exactly what SpliceCore is meant to supply.
Why it matters
The bet reflects two big trends at once: the industry’s hunt for fresh, tumor-selective targets to feed its powerful new delivery technologies, and the growing conviction that AI is most useful at the target-discovery end of the pipeline. If splicing-derived targets pan out, they could expand the universe of treatable cancers. “We hope to identify highly tumour-selective targets for the development of new cancer therapies to address the needs of patients who are still waiting for better treatment options,” Petronczki said.
The caveat
As with all such deals, the headline number is potential, not paid — most of the $1 billion is tied to milestones that may never be reached. This is an early-stage research partnership: identifying promising targets is a long way from a drug that works in patients, and computational predictions must be validated in the lab and then the clinic. The strategy is compelling; the proof will take years. Business news, not investment advice.