8f54

Publication Abstract from PubMed

As a result of evolutionary selection, the subunits of naturally occurring protein assemblies often fit together with substantial shape complementarity to generate architectures optimal for function in a manner not achievable by current design approaches. We describe a "top-down" reinforcement learning-based design approach that solves this problem using Monte Carlo tree search to sample protein conformers in the context of an overall architecture and specified functional constraints. Cryo-electron microscopy structures of the designed disk-shaped nanopores and ultracompact icosahedra are very close to the computational models. The icosohedra enable very-high-density display of immunogens and signaling molecules, which potentiates vaccine response and angiogenesis induction. Our approach enables the top-down design of complex protein nanomaterials with desired system properties and demonstrates the power of reinforcement learning in protein design.

Top-down design of protein architectures with reinforcement learning.,Lutz ID, Wang S, Norn C, Courbet A, Borst AJ, Zhao YT, Dosey A, Cao L, Xu J, Leaf EM, Treichel C, Litvicov P, Li Z, Goodson AD, Rivera-Sanchez P, Bratovianu AM, Baek M, King NP, Ruohola-Baker H, Baker D Science. 2023 Apr 21;380(6642):266-273. doi: 10.1126/science.adf6591. Epub 2023 , Apr 20. PMID:37079676^[1]

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.