In:
PLOS ONE, Public Library of Science (PLoS), Vol. 17, No. 3 ( 2022-3-14), p. e0265020-
Abstract:
Engineered proteins generally must possess a stable structure in order to achieve their designed function. Stable designs, however, are astronomically rare within the space of all possible amino acid sequences. As a consequence, many designs must be tested computationally and experimentally in order to find stable ones, which is expensive in terms of time and resources. Here we use a high-throughput, low-fidelity assay to experimentally evaluate the stability of approximately 200,000 novel proteins. These include a wide range of sequence perturbations, providing a baseline for future work in the field. We build a neural network model that predicts protein stability given only sequences of amino acids, and compare its performance to the assayed values. We also report another network model that is able to generate the amino acid sequences of novel stable proteins given requested secondary sequences. Finally, we show that the predictive model—despite weaknesses including a noisy data set—can be used to substantially increase the stability of both expert-designed and model-generated proteins.
Type of Medium:
Online Resource
ISSN:
1932-6203
DOI:
10.1371/journal.pone.0265020
DOI:
10.1371/journal.pone.0265020.g001
DOI:
10.1371/journal.pone.0265020.g002
DOI:
10.1371/journal.pone.0265020.g003
DOI:
10.1371/journal.pone.0265020.g004
DOI:
10.1371/journal.pone.0265020.g005
DOI:
10.1371/journal.pone.0265020.g006
DOI:
10.1371/journal.pone.0265020.g007
DOI:
10.1371/journal.pone.0265020.t001
DOI:
10.1371/journal.pone.0265020.t002
DOI:
10.1371/journal.pone.0265020.s001
DOI:
10.1371/journal.pone.0265020.s002
DOI:
10.1371/journal.pone.0265020.s003
DOI:
10.1371/journal.pone.0265020.s004
DOI:
10.1371/journal.pone.0265020.s005
DOI:
10.1371/journal.pone.0265020.s006
DOI:
10.1371/journal.pone.0265020.s007
DOI:
10.1371/journal.pone.0265020.s008
DOI:
10.1371/journal.pone.0265020.s009
DOI:
10.1371/journal.pone.0265020.s010
DOI:
10.1371/journal.pone.0265020.s011
DOI:
10.1371/journal.pone.0265020.s012
DOI:
10.1371/journal.pone.0265020.s013
DOI:
10.1371/journal.pone.0265020.s014
DOI:
10.1371/journal.pone.0265020.s015
DOI:
10.1371/journal.pone.0265020.s016
DOI:
10.1371/journal.pone.0265020.s017
DOI:
10.1371/journal.pone.0265020.s018
DOI:
10.1371/journal.pone.0265020.s019
DOI:
10.1371/journal.pone.0265020.s020
DOI:
10.1371/journal.pone.0265020.s021
DOI:
10.1371/journal.pone.0265020.s022
DOI:
10.1371/journal.pone.0265020.s023
DOI:
10.1371/journal.pone.0265020.s024
DOI:
10.1371/journal.pone.0265020.s025
DOI:
10.1371/journal.pone.0265020.s026
DOI:
10.1371/journal.pone.0265020.s027
DOI:
10.1371/journal.pone.0265020.s028
DOI:
10.1371/journal.pone.0265020.s029
DOI:
10.1371/journal.pone.0265020.s030
DOI:
10.1371/journal.pone.0265020.s031
DOI:
10.1371/journal.pone.0265020.s032
DOI:
10.1371/journal.pone.0265020.s033
DOI:
10.1371/journal.pone.0265020.s034
DOI:
10.1371/journal.pone.0265020.s035
DOI:
10.1371/journal.pone.0265020.s036
DOI:
10.1371/journal.pone.0265020.s037
DOI:
10.1371/journal.pone.0265020.s038
DOI:
10.1371/journal.pone.0265020.s039
DOI:
10.1371/journal.pone.0265020.s040
DOI:
10.1371/journal.pone.0265020.s041
DOI:
10.1371/journal.pone.0265020.r001
DOI:
10.1371/journal.pone.0265020.r002
DOI:
10.1371/journal.pone.0265020.r003
DOI:
10.1371/journal.pone.0265020.r004
DOI:
10.1371/journal.pone.0265020.r005
Language:
English
Publisher:
Public Library of Science (PLoS)
Publication Date:
2022
detail.hit.zdb_id:
2267670-3
Permalink