TY - JOUR
T1 - Robust deep learning-based protein sequence design using ProteinMPNN
AU - Dauparas, J.
AU - Anishchenko, I.
AU - Bennett, N.
AU - Bai, H.
AU - Ragotte, R.J.
AU - Milles, L.F.
AU - Wicky, B.I.M.
AU - Courbet, A.
AU - de Haas, R.J.
AU - Bethel, N.
AU - Leung, P.J.Y.
AU - Huddy, T.F.
AU - Pellock, S.
AU - Tischer, D.
AU - Chan, F.
AU - Koepnick, B.
AU - Nguyen, H.
AU - Kang, A.
AU - Sankaran, B.
AU - Bera, A.K.
AU - King, N.P.
AU - Baker, D.
PY - 2022/9/15
Y1 - 2022/9/15
N2 - Although deep learning has revolutionized protein structure prediction, almost all experimentally characterized de novo protein designs have been generated using physically based approaches such as Rosetta. Here, we describe a deep learning-based protein sequence design method, ProteinMPNN, that has outstanding performance in both in silico and experimental tests. On native protein backbones, ProteinMPNN has a sequence recovery of 52.4% compared with 32.9% for Rosetta. The amino acid sequence at different positions can be coupled between single or multiple chains, enabling application to a wide range of current protein design challenges. We demonstrate the broad utility and high accuracy of ProteinMPNN using x-ray crystallography, cryo-electron microscopy, and functional studies by rescuing previously failed designs, which were made using Rosetta or AlphaFold, of protein monomers, cyclic homo-oligomers, tetrahedral nanoparticles, and target-binding proteins.
AB - Although deep learning has revolutionized protein structure prediction, almost all experimentally characterized de novo protein designs have been generated using physically based approaches such as Rosetta. Here, we describe a deep learning-based protein sequence design method, ProteinMPNN, that has outstanding performance in both in silico and experimental tests. On native protein backbones, ProteinMPNN has a sequence recovery of 52.4% compared with 32.9% for Rosetta. The amino acid sequence at different positions can be coupled between single or multiple chains, enabling application to a wide range of current protein design challenges. We demonstrate the broad utility and high accuracy of ProteinMPNN using x-ray crystallography, cryo-electron microscopy, and functional studies by rescuing previously failed designs, which were made using Rosetta or AlphaFold, of protein monomers, cyclic homo-oligomers, tetrahedral nanoparticles, and target-binding proteins.
U2 - 10.1126/science.add2187
DO - 10.1126/science.add2187
M3 - Article
C2 - 36108050
AN - SCOPUS:85139380508
SN - 0036-8075
VL - 378
SP - 49
EP - 56
JO - Science (New York, N.Y.)
JF - Science (New York, N.Y.)
IS - 6615
ER -