MMDesign report finds experimental binders in 18 of 19 molecular design projects

What happened

QbitAI reports that 分子之心’s latest MMDesign technical report describes experimentally confirmed binders in 18 of 19 projects spanning 18 targets and two molecular modalities, VHH nanobodies and macrocyclic peptides. Each project tested 9–50 candidates; 139 of 453 tested candidates bound their targets, a 30.7% hit rate. All six tested GPCR targets yielded binders, with the best affinity for each reaching the nanomolar range. The report also describes improved results in five of six repeat design projects.

Read next

• More changes in today’s briefingTypeSafe AI and Treg release open-source people search tool powered by Jev

The story so far

Molecular Mind published a technical report on its MMDesign AI protein design platform, reporting experimentally confirmed binders in 18 of 19 projects, according to QbitAI. The projects covered 18 targets and two molecular modalities: VHH nanobodies and macrocyclic peptides. Each project tested 9–50 candidates. Across the projects, 139 of 453 tested candidates bound their targets, a 30.7% hit rate. All six tested GPCR targets yielded binders, with the best binder for each target achieving nanomolar affinity. The report also described improved results in five of six repeat design projects.