ai-drug-design
(7)
Ensem Therapeutics
Ensem Therapeutics is a Waltham, Massachusetts clinical-stage biotech using its Kinetic Ensemble platform - a blend of AI/ML, molecular simulation, and experimental biophysics - to find cryptic and allosteric pockets in proteins that are otherwise considered difficult to drug. The company designs small-molecule precision medicines for cancer, with two lead programs: ETX-636, an allosteric pan-mutant-selective PI3Kalpha inhibitor and degrader for breast cancer (FDA Fast Track), and ETX-197/BG-68501, a selective CDK2 inhibitor partnered with BeiGene.
Creyon Bio
Creyon Bio is a San Diego biotech reinventing how RNA-targeted medicines are designed. Founded in 2019 by former Ionis Pharmaceuticals scientists, the company pairs machine learning with quantum-level molecular modeling in its NucleIQ platform - an AI-Powered Oligo Engineering Engine - to predict the safety, efficacy, and delivery of oligonucleotide-based medicines before they ever reach a lab bench. The goal: replace slow, expensive trial-and-error with computational design rules that make precision therapeutics engineerable 'on demand,' for both rare and common diseases across neuromuscular, CNS, and immunologic areas.
Axelyf
Axelyf is a Brookline, Massachusetts biotech (with an R&D subsidiary in Iceland) building next-generation lipid nanoparticle delivery systems for RNA medicines. Founded by former Moderna drug-delivery leaders, it pairs a proprietary AXL ionizable-lipid library with an AI model called ANNA to make RNA and gene-editing payloads reach the right tissue with higher potency and less toxicity. In mouse liver studies its AXL LNPs showed roughly 4-5x higher gene-editing efficiency than the LP-01 clinical benchmark at equivalent doses, without evidence of liver toxicity. The company closed $2.6M in seed funding in August 2025 to push its lead autoimmune program and lipid library toward preclinical and non-human-primate studies.
Insamo
Insamo is a Berkeley-based biotechnology company spun out of UC Berkeley that is building membrane-permeable, orally available macrocyclic peptides with antibody-like binding affinity. By pairing machine-learning-driven molecular design with ultra-high-throughput molecular biology and parallel synthetic chemistry, Insamo runs automated design-synthesize-test cycles that generate trillions of proprietary data points - aiming to turn today's injected biologics into pills and to hit drug targets long considered undruggable. The company launched publicly in February 2024 with a $12M seed round led by Playground Global, venBio, and Merck's MRL Ventures Fund.
Vibrant Therapeutics
Vibrant Therapeutics (VibrantX) is a clinical-stage biotechnology company building intelligent, logic-gated antibody therapeutics that switch on only inside diseased tissue. By pairing computational protein design with high-throughput wet-lab validation across its BumbleBee, LogicBee and NeuroBee platforms, Vibrant develops conditionally activated, multi-specific antibody prodrugs for oncology, autoimmune, inflammatory and neurological diseases. Its lead program, VIB305, is a logic-gated masked T-cell engager for EGFR-positive solid tumors that has received FDA IND acceptance.
Totus Medicines
Totus Medicines is a clinical-stage precision medicines company in Emeryville, California, building covalent small-molecule drugs against historically undruggable targets. Its AI-powered OmniDEL platform screens billions of DNA-encoded covalent candidates against thousands of targets inside living cells, surfacing molecules that older methods miss. Its lead program, TOS-358, is the first and only covalent PI3Ka inhibitor in clinical development, showing class-leading tolerability and strong disease control in breast, endometrial, and head & neck cancers.
Fathom Therapeutics
Fathom Therapeutics (formerly Atommap) is a New York and Boston based biotech building Microcosmos, a drug-design engine that fuses quantum chemistry and AI to simulate how proteins actually move inside living cells. By modeling motion instead of static snapshots - reportedly up to 10,000x faster than conventional simulation - the company hunts for the fleeting binding pockets that make 'undruggable' targets druggable, then designs small molecules, degraders, and molecular glues against them. In April 2026 it raised a $47M oversubscribed Series A led by Sutter Hill Ventures.