AI in Healthcare

Eli Lilly Just Bet $2.25 Billion on AI-Designed Gene Editing Enzymes — Profluent's Recombinases Could Do What CRISPR Cannot

By Mustafa Shariff, Editor in Chief · May 11, 2026 · 2 min read

Summary

• Eli Lilly + Profluent: The $2.25 Billion AI Gene Editing Bet That Could Rewrite Genetic Medicine
• Eli Lilly has struck a deal worth up to $2.25 billion with AI biotech Profluent to develop AI-designed recombinases for genetic medicine. Here is the full story of what recombinases are and why this deal matters.

"Kilobase-scale DNA editing remains a holy grail in genetic medicine. Our work with Lilly is aimed at unlocking these therapeutics previously thought impossible. We believe only AI can create the designer recombinases needed to precisely target any location in the genome."

Eli Lilly — the pharmaceutical company whose GLP-1 obesity and diabetes drugs have generated record revenues and positioned it as one of the world's most valuable pharmaceutical companies — has entered a multi-programme research collaboration with Profluent, an Emeryville, California-based artificial intelligence biotech, to develop and commercialise custom site-specific recombinases for genetic diseases with severe unmet medical needs. The deal, announced on April 28, provides Profluent with an undisclosed upfront payment, committed R&D funding, and the opportunity to receive up to $2.25 billion in development and commercial milestone payments plus tiered royalties on net sales.

The collaboration is Lilly's second recombinase-focused transaction this year — and it comes alongside the company's earlier acquisition of Kelonia Therapeutics for up to $7 billion for in vivo CAR-T therapy, which BHC Media covered in our April 23 edition. Lilly is systematically constructing a genetic medicines portfolio built on AI-enabled approaches that go beyond the first generation of gene editing technologies.

Why Recombinases and Why Now

To understand the significance of this deal, it is necessary to understand what distinguishes recombinase-based gene editing from CRISPR — the gene editing technology that has dominated the genetic medicine landscape since Jennifer Doudna and Emmanuelle Charpentier's Nobel Prize-winning work. CRISPR uses a guide RNA and the Cas9 enzyme to cut DNA at a specific location and allows a repair template to insert or delete genetic material. It is highly effective for small edits — correcting single base pair mutations, disrupting specific genes, making targeted deletions. But CRISPR has a fundamental limitation: it struggles with large DNA insertions. Many genetic diseases require not a small correction but the insertion of an entire functional gene — a segment that may be thousands of base pairs in length, hence the term kilobase-scale editing.

Recombinases are a different class of enzyme. They work by recognising specific DNA sequences and catalysing a precise rearrangement — cutting and rejoining DNA at defined sites — in a way that is inherently more suited to large-scale DNA insertions. Natural recombinases found in bacteria and viruses operate with high fidelity but only recognise specific DNA sequences. The challenge for therapeutic applications is designing recombinases that recognise the specific genomic sequences relevant to human diseases — sequences that natural recombinases do not recognise.

Where AI Enters

This is precisely where Profluent's technology addresses a problem that conventional protein engineering cannot. The sequence space of possible recombinase variants is astronomically large — too large for conventional laboratory screening approaches to explore efficiently. Profluent applies large-scale foundation models for protein design — the same class of AI architecture that underlies large language models, but trained on protein sequence and structure data rather than text — to design custom recombinases that target specific genomic locations. The AI models can explore and evaluate millions of recombinase design candidates computationally, identifying variants with the targeting specificity and catalytic efficiency required for therapeutic use before any laboratory synthesis occurs.

Ali Madani, co-founder and CEO of Profluent, articulated the transformative aspiration: kilobase-scale DNA editing remains a holy grail in genetic medicine, and only AI can create the designer recombinases needed to precisely target any location in the genome. The goal is a fully programmable gene editing platform applicable across rare and common genetic diseases alike.

Lilly's Genetic Medicine Strategy

Lilly's investment in Profluent is part of a coherent genetic medicine strategy being executed with the company's Ozempic and Mounjaro-generated cash flows. Alongside the Profluent deal, Lilly has signed multiple genetic medicine partnerships targeting rare genetic diseases, haematological conditions, and oncology indications where conventional therapies are inadequate. The Kelonia in vivo CAR-T acquisition demonstrates that Lilly is simultaneously pursuing cell therapy and gene editing approaches — recognising that different disease categories may require different delivery mechanisms for genetic interventions.

Profluent is backed by investors including Eric Schmidt's family office, which reflects confidence in the AI-protein design approach from investors with deep exposure to AI capability assessments. The company has previously published open-source AI models for protein sequence generation, establishing scientific credibility in the AI-biology community that preceded the Lilly partnership.

What This Means for India

India carries a significant burden of genetic diseases — both common single-gene disorders like thalassaemia, haemophilia, and sickle cell disease, and rarer conditions that are nonetheless prevalent given India's population scale. The national thalassaemia burden alone affects an estimated 100,000 new patients per year. Sickle cell disease — for which Prime Minister Modi's government has launched the National Sickle Cell Anaemia Elimination Mission — affects millions of Indians, predominantly in tribal populations. Genetic medicine breakthroughs that emerge from Lilly-Profluent collaboration will eventually create treatment options for these populations, though the timeline from research to Indian market access remains long.

BHC Perspective

At Bengaluru Health Community, the Lilly-Profluent deal is the most significant signal yet that AI-designed proteins are moving from academic research into the commercial pharmaceutical development pipeline at serious scale. For BHC founders building at the intersection of AI and biology — drug discovery, protein engineering, genomics platforms — this $2.25 billion bet by the world's most commercially successful pharmaceutical company validates the investment thesis in the clearest possible terms.

Frequently Asked Questions

Q: What is the Profluent-Eli Lilly deal?
A: A multi-programme research collaboration announced April 28 in which Profluent will use AI models to design custom site-specific recombinases for multiple genomic targets, with Lilly receiving exclusive licences to advance selected recombinases into preclinical and clinical development. Profluent is eligible for up to $2.25 billion in milestones plus royalties.

Q: What are recombinases and why do they matter for genetic medicine?
A: Recombinases are enzymes that cut and rejoin DNA at specific sequences. Unlike CRISPR, which struggles with large DNA insertions, engineered recombinases can enable kilobase-scale DNA editing — inserting entire functional genes — making them potentially therapeutic for diseases requiring large genetic corrections.

Q: Why does this deal require AI?
A: The sequence space of possible recombinase variants is too large for conventional laboratory screening. Profluent uses large-scale AI foundation models trained on protein sequence and structure data to design custom recombinases targeting specific human genomic sequences — enabling therapeutic use at precision and scale impossible with conventional protein engineering.