Evaluation of cross-validation strategies in sequence-based binding prediction using deep learning

Binding prediction between targets and drug-like compounds through deep neural networks has generated promising results in recent years, outperforming traditional machine learning-based methods. However, the generalization capability of these classification models is still an issue to be addressed....

Full description

Bibliographic Details
Authors: López del Río, Ángela|||0000-0002-5486-7465, Nonell Canals, Alfons, Vidal, David, Perera Lluna, Alexandre|||0000-0001-6427-851X
Format: article
Publication Date:2019
Country:España
Institution:Universitat Politècnica de Catalunya (UPC)
Repository:UPCommons. Portal del coneixement obert de la UPC
Language:English
OAI Identifier:oai:upcommons.upc.edu:2117/168430
Online Access:https://hdl.handle.net/2117/168430
https://dx.doi.org/10.1021/acs.jcim.8b00663
Access Level:Open access
Keyword:Deep neural networks
Machine learning
Aprenentatge automàtic
Àrees temàtiques de la UPC::Informàtica::Aplicacions de la informàtica::Aplicacions informàtiques a la física i l‘enginyeria
id ES_818bffb2ee1448748cf31fc6ec2986bb
oai_identifier_str oai:upcommons.upc.edu:2117/168430
network_acronym_str ES
network_name_str España
repository_id_str
spelling Evaluation of cross-validation strategies in sequence-based binding prediction using deep learningLópez del Río, Ángela|||0000-0002-5486-7465Nonell Canals, AlfonsVidal, DavidPerera Lluna, Alexandre|||0000-0001-6427-851XDeep neural networksMachine learningAprenentatge automàticÀrees temàtiques de la UPC::Informàtica::Aplicacions de la informàtica::Aplicacions informàtiques a la física i l‘enginyeriaBinding prediction between targets and drug-like compounds through deep neural networks has generated promising results in recent years, outperforming traditional machine learning-based methods. However, the generalization capability of these classification models is still an issue to be addressed. In this work, we explored how different cross-validation strategies applied to data from different molecular databases affect to the performance of binding prediction proteochemometrics models. These strategies are (1) random splitting, (2) splitting based on K-means clustering (both of actives and inactives), (3) splitting based on source database, and (4) splitting based both in the clustering and in the source database. These schemas are applied to a deep learning proteochemometrics model and to a simple logistic regression model to be used as baseline. Additionally, two different ways of describing molecules in the model are tested: (1) by their SMILES and (2) by three fingerprints. The classification performance of our deep learning-based proteochemometrics model is comparable to the state of the art. Our results show that the lack of generalization of these models is due to a bias in public molecular databases and that a restrictive cross-validation schema based on compound clustering leads to worse but more robust and credible results. Our results also show better performance when representing molecules by their fingerprints.Peer Reviewed20192019-01-0120192019-09-19journal articlehttp://purl.org/coar/resource_type/c_6501AMhttp://purl.org/coar/version/c_ab4af688f83e57aainfo:eu-repo/semantics/articleapplication/pdfhttps://hdl.handle.net/2117/168430https://dx.doi.org/10.1021/acs.jcim.8b00663reponame:UPCommons. Portal del coneixement obert de la UPCinstname:Universitat Politècnica de Catalunya (UPC)Inglésengopen accesshttp://purl.org/coar/access_right/c_abf2info:eu-repo/semantics/openAccessoai:upcommons.upc.edu:2117/1684302026-05-27T15:37:01Z
dc.title.none.fl_str_mv Evaluation of cross-validation strategies in sequence-based binding prediction using deep learning
title Evaluation of cross-validation strategies in sequence-based binding prediction using deep learning
spellingShingle Evaluation of cross-validation strategies in sequence-based binding prediction using deep learning
López del Río, Ángela|||0000-0002-5486-7465
Deep neural networks
Machine learning
Aprenentatge automàtic
Àrees temàtiques de la UPC::Informàtica::Aplicacions de la informàtica::Aplicacions informàtiques a la física i l‘enginyeria
title_short Evaluation of cross-validation strategies in sequence-based binding prediction using deep learning
title_full Evaluation of cross-validation strategies in sequence-based binding prediction using deep learning
title_fullStr Evaluation of cross-validation strategies in sequence-based binding prediction using deep learning
title_full_unstemmed Evaluation of cross-validation strategies in sequence-based binding prediction using deep learning
title_sort Evaluation of cross-validation strategies in sequence-based binding prediction using deep learning
dc.creator.none.fl_str_mv López del Río, Ángela|||0000-0002-5486-7465
Nonell Canals, Alfons
Vidal, David
Perera Lluna, Alexandre|||0000-0001-6427-851X
author López del Río, Ángela|||0000-0002-5486-7465
author_facet López del Río, Ángela|||0000-0002-5486-7465
Nonell Canals, Alfons
Vidal, David
Perera Lluna, Alexandre|||0000-0001-6427-851X
author_role author
author2 Nonell Canals, Alfons
Vidal, David
Perera Lluna, Alexandre|||0000-0001-6427-851X
author2_role author
author
author
dc.subject.none.fl_str_mv Deep neural networks
Machine learning
Aprenentatge automàtic
Àrees temàtiques de la UPC::Informàtica::Aplicacions de la informàtica::Aplicacions informàtiques a la física i l‘enginyeria
topic Deep neural networks
Machine learning
Aprenentatge automàtic
Àrees temàtiques de la UPC::Informàtica::Aplicacions de la informàtica::Aplicacions informàtiques a la física i l‘enginyeria
description Binding prediction between targets and drug-like compounds through deep neural networks has generated promising results in recent years, outperforming traditional machine learning-based methods. However, the generalization capability of these classification models is still an issue to be addressed. In this work, we explored how different cross-validation strategies applied to data from different molecular databases affect to the performance of binding prediction proteochemometrics models. These strategies are (1) random splitting, (2) splitting based on K-means clustering (both of actives and inactives), (3) splitting based on source database, and (4) splitting based both in the clustering and in the source database. These schemas are applied to a deep learning proteochemometrics model and to a simple logistic regression model to be used as baseline. Additionally, two different ways of describing molecules in the model are tested: (1) by their SMILES and (2) by three fingerprints. The classification performance of our deep learning-based proteochemometrics model is comparable to the state of the art. Our results show that the lack of generalization of these models is due to a bias in public molecular databases and that a restrictive cross-validation schema based on compound clustering leads to worse but more robust and credible results. Our results also show better performance when representing molecules by their fingerprints.
publishDate 2019
dc.date.none.fl_str_mv 2019
2019-01-01
2019
2019-09-19
dc.type.none.fl_str_mv journal article
http://purl.org/coar/resource_type/c_6501
AM
http://purl.org/coar/version/c_ab4af688f83e57aa
dc.type.openaire.fl_str_mv info:eu-repo/semantics/article
format article
dc.identifier.none.fl_str_mv https://hdl.handle.net/2117/168430
https://dx.doi.org/10.1021/acs.jcim.8b00663
url https://hdl.handle.net/2117/168430
https://dx.doi.org/10.1021/acs.jcim.8b00663
dc.language.none.fl_str_mv Inglés
eng
language_invalid_str_mv Inglés
language eng
dc.rights.none.fl_str_mv open access
http://purl.org/coar/access_right/c_abf2
dc.rights.openaire.fl_str_mv info:eu-repo/semantics/openAccess
rights_invalid_str_mv open access
http://purl.org/coar/access_right/c_abf2
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.source.none.fl_str_mv reponame:UPCommons. Portal del coneixement obert de la UPC
instname:Universitat Politècnica de Catalunya (UPC)
instname_str Universitat Politècnica de Catalunya (UPC)
reponame_str UPCommons. Portal del coneixement obert de la UPC
collection UPCommons. Portal del coneixement obert de la UPC
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869411981509787648
score 15.301603

Similar Items