Loading
ESM3 is a generative protein model that designed a working fluorescent protein far from any natural sequence.
ESM3 from EvolutionaryScale is a generative protein language model that jointly represents sequence, structure and function as tokens in a masked multimodal transformer, so a user can prompt it with any combination of these and ask it to fill in the rest. The Science 2025 paper describes models up to 98B parameters and the design of esmGFP, a working fluorescent protein far from any natural sequence, demonstrating generation of novel functional proteins. It is aimed at protein engineers and drug designers exploring binders, enzymes and antibodies, including for cancer targets. The largest weights are closed, with smaller versions open, so the strongest capabilities are available only through the company. For a newcomer: ESM3 has learned enough about proteins to invent new ones that work.
ESM3 is a masked multimodal transformer over protein tokens.
Query for this technology: (TITLE:"ESM3" OR ABSTRACT:"ESM3" OR TITLE:"EvolutionaryScale" OR ABSTRACT:"EvolutionaryScale") AND (cancer OR tumor OR tumour OR oncology OR carcinoma OR lymphoma OR leukemia OR leukaemia OR myeloma OR sarcoma OR melanoma OR glioma). Results are unfiltered search hits about ESM3 (EvolutionaryScale), not a curated reading list.
Shares Drug discovery roadmap: screening in mice → maps of dependency → designing in silico, AI in oncology roadmap: pattern readers → foundation models → agents in the workflow, AI-driven drug & target discovery and the tags foundation-model, protein-design.
Shares Drug discovery roadmap: screening in mice → maps of dependency → designing in silico, AI in oncology roadmap: pattern readers → foundation models → agents in the workflow, AI-driven drug & target discovery and the tags foundation-model, protein-design.
Shares Drug discovery roadmap: screening in mice → maps of dependency → designing in silico, AI in oncology roadmap: pattern readers → foundation models → agents in the workflow, AI-driven drug & target discovery and the tag foundation-model.
Shares Drug discovery roadmap: screening in mice → maps of dependency → designing in silico, AI in oncology roadmap: pattern readers → foundation models → agents in the workflow, AI-driven drug & target discovery and the tag foundation-model.
Shares Drug discovery roadmap: screening in mice → maps of dependency → designing in silico, AI in oncology roadmap: pattern readers → foundation models → agents in the workflow, AI-driven drug & target discovery and the tag foundation-model.
Shares Drug discovery roadmap: screening in mice → maps of dependency → designing in silico, AI in oncology roadmap: pattern readers → foundation models → agents in the workflow and the tag foundation-model.
Shares Drug discovery roadmap: screening in mice → maps of dependency → designing in silico, AI in oncology roadmap: pattern readers → foundation models → agents in the workflow and the tag foundation-model.
Shares Drug discovery roadmap: screening in mice → maps of dependency → designing in silico, AI in oncology roadmap: pattern readers → foundation models → agents in the workflow and the tag foundation-model.
Open-source projects that implement or serve this technology, from OnCo's own catalogue: licence and last activity as the repository reported them on the day of the fetch. Listing is not endorsement; check the licence before reuse and the validation before clinical use.
EvolutionaryScale's protein language models; small models have open weights, larger ones are under a non-commercial licence.