Cargando…

Construction of a Deep Neural Network Energy Function for Protein Physics

[Image: see text] The traditional approach of computational biology consists of calculating molecule properties by using approximate classical potentials. Interactions between atoms are described by an energy function derived from physical principles or fitted to experimental data. Their functional...

Descripción completa

Detalles Bibliográficos
Autores principales: Yang, Huan, Xiong, Zhaoping, Zonta, Francesco
Formato: Online Artículo Texto
Lenguaje:English
Publicado: American Chemical Society 2022
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9476656/
https://www.ncbi.nlm.nih.gov/pubmed/35939398
http://dx.doi.org/10.1021/acs.jctc.2c00069
_version_ 1784790186255712256
author Yang, Huan
Xiong, Zhaoping
Zonta, Francesco
author_facet Yang, Huan
Xiong, Zhaoping
Zonta, Francesco
author_sort Yang, Huan
collection PubMed
description [Image: see text] The traditional approach of computational biology consists of calculating molecule properties by using approximate classical potentials. Interactions between atoms are described by an energy function derived from physical principles or fitted to experimental data. Their functional form is usually limited to pairwise interactions between atoms and does not consider complex multibody effects. More recently, neural networks have emerged as an alternative way of describing the interactions between biomolecules. In this approach, the energy function does not have an explicit functional form and is learned bottom-up from simulations at the atomistic or quantum level. In this study, we attempt a top-down approach and use deep learning methods to obtain an energy function by exploiting the large amount of experimental data acquired with years in the field of structural biology. The energy function is represented by a probability density model learned from a large repertoire of building blocks representing local clusters of amino acids paired with their sequence signature. We demonstrated the feasibility of this approach by generating a neural network energy function and testing its validity on several applications such as discriminating decoys, assessing qualities of structural models, sampling structural conformations, and designing new protein sequences. We foresee that, in the future, our methodology could exploit the continuously increasing availability of experimental data and simulations and provide a new method for the parametrization of protein energy functions.
format Online
Article
Text
id pubmed-9476656
institution National Center for Biotechnology Information
language English
publishDate 2022
publisher American Chemical Society
record_format MEDLINE/PubMed
spelling pubmed-94766562022-09-16 Construction of a Deep Neural Network Energy Function for Protein Physics Yang, Huan Xiong, Zhaoping Zonta, Francesco J Chem Theory Comput [Image: see text] The traditional approach of computational biology consists of calculating molecule properties by using approximate classical potentials. Interactions between atoms are described by an energy function derived from physical principles or fitted to experimental data. Their functional form is usually limited to pairwise interactions between atoms and does not consider complex multibody effects. More recently, neural networks have emerged as an alternative way of describing the interactions between biomolecules. In this approach, the energy function does not have an explicit functional form and is learned bottom-up from simulations at the atomistic or quantum level. In this study, we attempt a top-down approach and use deep learning methods to obtain an energy function by exploiting the large amount of experimental data acquired with years in the field of structural biology. The energy function is represented by a probability density model learned from a large repertoire of building blocks representing local clusters of amino acids paired with their sequence signature. We demonstrated the feasibility of this approach by generating a neural network energy function and testing its validity on several applications such as discriminating decoys, assessing qualities of structural models, sampling structural conformations, and designing new protein sequences. We foresee that, in the future, our methodology could exploit the continuously increasing availability of experimental data and simulations and provide a new method for the parametrization of protein energy functions. American Chemical Society 2022-08-08 2022-09-13 /pmc/articles/PMC9476656/ /pubmed/35939398 http://dx.doi.org/10.1021/acs.jctc.2c00069 Text en © 2022 The Authors. Published by American Chemical Society https://creativecommons.org/licenses/by-nc-nd/4.0/Permits non-commercial access and re-use, provided that author attribution and integrity are maintained; but does not permit creation of adaptations or other derivative works (https://creativecommons.org/licenses/by-nc-nd/4.0/).
spellingShingle Yang, Huan
Xiong, Zhaoping
Zonta, Francesco
Construction of a Deep Neural Network Energy Function for Protein Physics
title Construction of a Deep Neural Network Energy Function for Protein Physics
title_full Construction of a Deep Neural Network Energy Function for Protein Physics
title_fullStr Construction of a Deep Neural Network Energy Function for Protein Physics
title_full_unstemmed Construction of a Deep Neural Network Energy Function for Protein Physics
title_short Construction of a Deep Neural Network Energy Function for Protein Physics
title_sort construction of a deep neural network energy function for protein physics
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9476656/
https://www.ncbi.nlm.nih.gov/pubmed/35939398
http://dx.doi.org/10.1021/acs.jctc.2c00069
work_keys_str_mv AT yanghuan constructionofadeepneuralnetworkenergyfunctionforproteinphysics
AT xiongzhaoping constructionofadeepneuralnetworkenergyfunctionforproteinphysics
AT zontafrancesco constructionofadeepneuralnetworkenergyfunctionforproteinphysics