Cargando…

Does a More Precise Chemical Description of Protein–Ligand Complexes Lead to More Accurate Prediction of Binding Affinity?

[Image: see text] Predicting the binding affinities of large sets of diverse molecules against a range of macromolecular targets is an extremely challenging task. The scoring functions that attempt such computational prediction are essential for exploiting and analyzing the outputs of docking, which...

Descripción completa

Detalles Bibliográficos
Autores principales: Ballester, Pedro J., Schreyer, Adrian, Blundell, Tom L.
Formato: Online Artículo Texto
Lenguaje:English
Publicado: American Chemical Society 2014
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3966527/
https://www.ncbi.nlm.nih.gov/pubmed/24528282
http://dx.doi.org/10.1021/ci500091r
_version_ 1782308924476620800
author Ballester, Pedro J.
Schreyer, Adrian
Blundell, Tom L.
author_facet Ballester, Pedro J.
Schreyer, Adrian
Blundell, Tom L.
author_sort Ballester, Pedro J.
collection PubMed
description [Image: see text] Predicting the binding affinities of large sets of diverse molecules against a range of macromolecular targets is an extremely challenging task. The scoring functions that attempt such computational prediction are essential for exploiting and analyzing the outputs of docking, which is in turn an important tool in problems such as structure-based drug design. Classical scoring functions assume a predetermined theory-inspired functional form for the relationship between the variables that describe an experimentally determined or modeled structure of a protein–ligand complex and its binding affinity. The inherent problem of this approach is in the difficulty of explicitly modeling the various contributions of intermolecular interactions to binding affinity. New scoring functions based on machine-learning regression models, which are able to exploit effectively much larger amounts of experimental data and circumvent the need for a predetermined functional form, have already been shown to outperform a broad range of state-of-the-art scoring functions in a widely used benchmark. Here, we investigate the impact of the chemical description of the complex on the predictive power of the resulting scoring function using a systematic battery of numerical experiments. The latter resulted in the most accurate scoring function to date on the benchmark. Strikingly, we also found that a more precise chemical description of the protein–ligand complex does not generally lead to a more accurate prediction of binding affinity. We discuss four factors that may contribute to this result: modeling assumptions, codependence of representation and regression, data restricted to the bound state, and conformational heterogeneity in data.
format Online
Article
Text
id pubmed-3966527
institution National Center for Biotechnology Information
language English
publishDate 2014
publisher American Chemical Society
record_format MEDLINE/PubMed
spelling pubmed-39665272014-03-27 Does a More Precise Chemical Description of Protein–Ligand Complexes Lead to More Accurate Prediction of Binding Affinity? Ballester, Pedro J. Schreyer, Adrian Blundell, Tom L. J Chem Inf Model [Image: see text] Predicting the binding affinities of large sets of diverse molecules against a range of macromolecular targets is an extremely challenging task. The scoring functions that attempt such computational prediction are essential for exploiting and analyzing the outputs of docking, which is in turn an important tool in problems such as structure-based drug design. Classical scoring functions assume a predetermined theory-inspired functional form for the relationship between the variables that describe an experimentally determined or modeled structure of a protein–ligand complex and its binding affinity. The inherent problem of this approach is in the difficulty of explicitly modeling the various contributions of intermolecular interactions to binding affinity. New scoring functions based on machine-learning regression models, which are able to exploit effectively much larger amounts of experimental data and circumvent the need for a predetermined functional form, have already been shown to outperform a broad range of state-of-the-art scoring functions in a widely used benchmark. Here, we investigate the impact of the chemical description of the complex on the predictive power of the resulting scoring function using a systematic battery of numerical experiments. The latter resulted in the most accurate scoring function to date on the benchmark. Strikingly, we also found that a more precise chemical description of the protein–ligand complex does not generally lead to a more accurate prediction of binding affinity. We discuss four factors that may contribute to this result: modeling assumptions, codependence of representation and regression, data restricted to the bound state, and conformational heterogeneity in data. American Chemical Society 2014-02-16 2014-03-24 /pmc/articles/PMC3966527/ /pubmed/24528282 http://dx.doi.org/10.1021/ci500091r Text en Copyright © 2014 American Chemical Society Terms of Use (http://pubs.acs.org/page/policy/authorchoice_termsofuse.html)
spellingShingle Ballester, Pedro J.
Schreyer, Adrian
Blundell, Tom L.
Does a More Precise Chemical Description of Protein–Ligand Complexes Lead to More Accurate Prediction of Binding Affinity?
title Does a More Precise Chemical Description of Protein–Ligand Complexes Lead to More Accurate Prediction of Binding Affinity?
title_full Does a More Precise Chemical Description of Protein–Ligand Complexes Lead to More Accurate Prediction of Binding Affinity?
title_fullStr Does a More Precise Chemical Description of Protein–Ligand Complexes Lead to More Accurate Prediction of Binding Affinity?
title_full_unstemmed Does a More Precise Chemical Description of Protein–Ligand Complexes Lead to More Accurate Prediction of Binding Affinity?
title_short Does a More Precise Chemical Description of Protein–Ligand Complexes Lead to More Accurate Prediction of Binding Affinity?
title_sort does a more precise chemical description of protein–ligand complexes lead to more accurate prediction of binding affinity?
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3966527/
https://www.ncbi.nlm.nih.gov/pubmed/24528282
http://dx.doi.org/10.1021/ci500091r
work_keys_str_mv AT ballesterpedroj doesamoreprecisechemicaldescriptionofproteinligandcomplexesleadtomoreaccuratepredictionofbindingaffinity
AT schreyeradrian doesamoreprecisechemicaldescriptionofproteinligandcomplexesleadtomoreaccuratepredictionofbindingaffinity
AT blundelltoml doesamoreprecisechemicaldescriptionofproteinligandcomplexesleadtomoreaccuratepredictionofbindingaffinity