DeepDDG: Predicting the Stability Change of Protein Point Mutations Using Neural Networks

Huali Cao, Jingxue Wang, Liping He, Yifei Qi, John Zhang

Research output: Contribution to journalArticle

Abstract

Accurately predicting changes in protein stability due to mutations is important for protein engineering and for understanding the functional consequences of missense mutations in proteins. We have developed DeepDDG, a neural network-based method, for use in the prediction of changes in the stability of proteins due to point mutations. The neural network was trained on more than 5700 manually curated experimental data points and was able to obtain a Pearson correlation coefficient of 0.48-0.56 for three independent test sets, which outperformed 11 other methods. Detailed analysis of the input features shows that the solvent accessible surface area of the mutated residue is the most important feature, which suggests that the buried hydrophobic area is the major determinant of protein stability. We expect this method to be useful for large-scale design and engineering of protein stability. The neural network is freely available to academic users at http://protein.org.cn/ddg.html.

Original languageEnglish (US)
Pages (from-to)1508-1514
Number of pages7
JournalJournal of Chemical Information and Modeling
Volume59
Issue number4
DOIs
StatePublished - Apr 22 2019

Fingerprint

neural network
Neural networks
Proteins
engineering
determinants

ASJC Scopus subject areas

  • Chemistry(all)
  • Chemical Engineering(all)
  • Computer Science Applications
  • Library and Information Sciences

Cite this

DeepDDG : Predicting the Stability Change of Protein Point Mutations Using Neural Networks. / Cao, Huali; Wang, Jingxue; He, Liping; Qi, Yifei; Zhang, John.

In: Journal of Chemical Information and Modeling, Vol. 59, No. 4, 22.04.2019, p. 1508-1514.

Research output: Contribution to journalArticle

Cao, Huali ; Wang, Jingxue ; He, Liping ; Qi, Yifei ; Zhang, John. / DeepDDG : Predicting the Stability Change of Protein Point Mutations Using Neural Networks. In: Journal of Chemical Information and Modeling. 2019 ; Vol. 59, No. 4. pp. 1508-1514.
@article{a9efc5fa15d549af9310849f159b800e,
title = "DeepDDG: Predicting the Stability Change of Protein Point Mutations Using Neural Networks",
abstract = "Accurately predicting changes in protein stability due to mutations is important for protein engineering and for understanding the functional consequences of missense mutations in proteins. We have developed DeepDDG, a neural network-based method, for use in the prediction of changes in the stability of proteins due to point mutations. The neural network was trained on more than 5700 manually curated experimental data points and was able to obtain a Pearson correlation coefficient of 0.48-0.56 for three independent test sets, which outperformed 11 other methods. Detailed analysis of the input features shows that the solvent accessible surface area of the mutated residue is the most important feature, which suggests that the buried hydrophobic area is the major determinant of protein stability. We expect this method to be useful for large-scale design and engineering of protein stability. The neural network is freely available to academic users at http://protein.org.cn/ddg.html.",
author = "Huali Cao and Jingxue Wang and Liping He and Yifei Qi and John Zhang",
year = "2019",
month = "4",
day = "22",
doi = "10.1021/acs.jcim.8b00697",
language = "English (US)",
volume = "59",
pages = "1508--1514",
journal = "Journal of Chemical Information and Modeling",
issn = "1549-9596",
publisher = "American Chemical Society",
number = "4",

}

TY - JOUR

T1 - DeepDDG

T2 - Predicting the Stability Change of Protein Point Mutations Using Neural Networks

AU - Cao, Huali

AU - Wang, Jingxue

AU - He, Liping

AU - Qi, Yifei

AU - Zhang, John

PY - 2019/4/22

Y1 - 2019/4/22

N2 - Accurately predicting changes in protein stability due to mutations is important for protein engineering and for understanding the functional consequences of missense mutations in proteins. We have developed DeepDDG, a neural network-based method, for use in the prediction of changes in the stability of proteins due to point mutations. The neural network was trained on more than 5700 manually curated experimental data points and was able to obtain a Pearson correlation coefficient of 0.48-0.56 for three independent test sets, which outperformed 11 other methods. Detailed analysis of the input features shows that the solvent accessible surface area of the mutated residue is the most important feature, which suggests that the buried hydrophobic area is the major determinant of protein stability. We expect this method to be useful for large-scale design and engineering of protein stability. The neural network is freely available to academic users at http://protein.org.cn/ddg.html.

AB - Accurately predicting changes in protein stability due to mutations is important for protein engineering and for understanding the functional consequences of missense mutations in proteins. We have developed DeepDDG, a neural network-based method, for use in the prediction of changes in the stability of proteins due to point mutations. The neural network was trained on more than 5700 manually curated experimental data points and was able to obtain a Pearson correlation coefficient of 0.48-0.56 for three independent test sets, which outperformed 11 other methods. Detailed analysis of the input features shows that the solvent accessible surface area of the mutated residue is the most important feature, which suggests that the buried hydrophobic area is the major determinant of protein stability. We expect this method to be useful for large-scale design and engineering of protein stability. The neural network is freely available to academic users at http://protein.org.cn/ddg.html.

UR - http://www.scopus.com/inward/record.url?scp=85064650136&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85064650136&partnerID=8YFLogxK

U2 - 10.1021/acs.jcim.8b00697

DO - 10.1021/acs.jcim.8b00697

M3 - Article

C2 - 30759982

AN - SCOPUS:85064650136

VL - 59

SP - 1508

EP - 1514

JO - Journal of Chemical Information and Modeling

JF - Journal of Chemical Information and Modeling

SN - 1549-9596

IS - 4

ER -