Abstract
The task of protein sequence design is central to nearly all rational protein engineering problems, and enormous effort has gone into the development of energy functions to guide design. We investigate the capability of a deep neural network model to automate design of sequences onto protein backbones, having learned directly from crystal structure data and without any human-specified priors. The model generalizes to native topologies not seen during training, producing experimentally stable designs. We evaluate the generalizability of our method to a de novo TIM-barrel scaffold. The model produces novel sequences, and high-resolution crystal structures of two designs show excellent agreement with the in silico models. Our findings demonstrate the tractability of an entirely learned method for protein sequence design.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
namrataa{at}stanford.edu, reguchi{at}stanford.edu, iimathew{at}slac.stanford.edu, cperez8{at}stanford.edu, aderry{at}stanford.edu, russ.altman{at}stanford.edu, possu{at}stanford.edu
New experimental validation results
https://drive.google.com/drive/folders/1gBfu5LG8-kp9o7qBMkCdBSRCd0E8R6Te?usp=sharing
