A Ligand-Based Approach to the Discovery of Lead-Like Potassium Channel KV1.3 Inhibitors

Gilberto Pereira, Beatriz Szwarc, Miguel A. Mondragão, Pedro A. Lima, Florbela Pereira

Research output: Contribution to journalArticle

Abstract

Voltage-gated ion channels are key molecular targets for autoimmune diseases such as multiple sclerosis, rheumatoid arthritis and psoriasis. In silico models, using 340 molecules whose IC50 towards Kv1.3 was determined by reported assays, were developed through exploration of four machine learning (ML) techniques. ML techniques explored included Random Forest, Support Vector Machine, Multilayer Perceptron, and K-Nearest Neighbors. Two QSAR classification approaches were developed. In the first approach, the compounds were classified into either moderate-active-to-very-active or inactive-to-moderate-active categories. Only the compounds predicted to be moderate-active-to-very-active in the first classification model were submitted to the second model, a classification model that predicted two more categories, very-active and not-very-active. The performances of the models were successfully evaluated by internal validation and external test set validation, with an overall predictability (Q) of 0.83 and 0.69 for the test set in the first and second approaches of the best models, respectively. The best models for the two approaches were employed for the virtual screening of KV1.3 inhibitors from ZINC natural products and approved drugs databases. A list of the most promising lead-like KV1.3 inhibitors was proposed, from which an approved drug and a natural product were experimentally evaluated with whole-cell voltage-clamp assays at 1000 nM.

Original languageEnglish
Pages (from-to)1352-1364
Number of pages13
JournalChemistrySelect
Volume3
Issue number5
DOIs
Publication statusPublished - 7 Feb 2018

Fingerprint

Kv1.3 Potassium Channel
Ligands
Biological Products
Learning systems
Assays
Clamping devices
Electric potential
Multilayer neural networks
Lead
Ion Channels
Pharmaceutical Preparations
Support vector machines
Screening

Keywords

  • K1.3 inhibitors
  • machine learning (ML)
  • quantitative structure activity relationships (QSAR)
  • random forest (RF)
  • virtual screening

Cite this

Pereira, Gilberto ; Szwarc, Beatriz ; Mondragão, Miguel A. ; Lima, Pedro A. ; Pereira, Florbela. / A Ligand-Based Approach to the Discovery of Lead-Like Potassium Channel KV1.3 Inhibitors. In: ChemistrySelect. 2018 ; Vol. 3, No. 5. pp. 1352-1364.
@article{8e72e9a3c8bf4372b6f395ca37f488bf,
title = "A Ligand-Based Approach to the Discovery of Lead-Like Potassium Channel KV1.3 Inhibitors",
abstract = "Voltage-gated ion channels are key molecular targets for autoimmune diseases such as multiple sclerosis, rheumatoid arthritis and psoriasis. In silico models, using 340 molecules whose IC50 towards Kv1.3 was determined by reported assays, were developed through exploration of four machine learning (ML) techniques. ML techniques explored included Random Forest, Support Vector Machine, Multilayer Perceptron, and K-Nearest Neighbors. Two QSAR classification approaches were developed. In the first approach, the compounds were classified into either moderate-active-to-very-active or inactive-to-moderate-active categories. Only the compounds predicted to be moderate-active-to-very-active in the first classification model were submitted to the second model, a classification model that predicted two more categories, very-active and not-very-active. The performances of the models were successfully evaluated by internal validation and external test set validation, with an overall predictability (Q) of 0.83 and 0.69 for the test set in the first and second approaches of the best models, respectively. The best models for the two approaches were employed for the virtual screening of KV1.3 inhibitors from ZINC natural products and approved drugs databases. A list of the most promising lead-like KV1.3 inhibitors was proposed, from which an approved drug and a natural product were experimentally evaluated with whole-cell voltage-clamp assays at 1000 nM.",
keywords = "K1.3 inhibitors, machine learning (ML), quantitative structure activity relationships (QSAR), random forest (RF), virtual screening",
author = "Gilberto Pereira and Beatriz Szwarc and Mondrag{\~a}o, {Miguel A.} and Lima, {Pedro A.} and Florbela Pereira",
note = "info:eu-repo/grantAgreement/FCT/3599-PPCDT/132949/PT# info:eu-repo/grantAgreement/FCT/5876/147218/PT# sem pdf conforme despacho. Funda{\cc}{\~a}o para a Ci{\^e}ncia e a Tecnologia (FCT/MEC) , SFRH/BPD/108237/2015. PT2020. Grant Number: POCI-01-0145-FEDER-007265",
year = "2018",
month = "2",
day = "7",
doi = "10.1002/slct.201702977",
language = "English",
volume = "3",
pages = "1352--1364",
journal = "ChemistrySelect",
issn = "2365-6549",
publisher = "Wiley",
number = "5",

}

A Ligand-Based Approach to the Discovery of Lead-Like Potassium Channel KV1.3 Inhibitors. / Pereira, Gilberto; Szwarc, Beatriz; Mondragão, Miguel A.; Lima, Pedro A.; Pereira, Florbela.

In: ChemistrySelect, Vol. 3, No. 5, 07.02.2018, p. 1352-1364.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A Ligand-Based Approach to the Discovery of Lead-Like Potassium Channel KV1.3 Inhibitors

AU - Pereira, Gilberto

AU - Szwarc, Beatriz

AU - Mondragão, Miguel A.

AU - Lima, Pedro A.

AU - Pereira, Florbela

N1 - info:eu-repo/grantAgreement/FCT/3599-PPCDT/132949/PT# info:eu-repo/grantAgreement/FCT/5876/147218/PT# sem pdf conforme despacho. Fundação para a Ciência e a Tecnologia (FCT/MEC) , SFRH/BPD/108237/2015. PT2020. Grant Number: POCI-01-0145-FEDER-007265

PY - 2018/2/7

Y1 - 2018/2/7

N2 - Voltage-gated ion channels are key molecular targets for autoimmune diseases such as multiple sclerosis, rheumatoid arthritis and psoriasis. In silico models, using 340 molecules whose IC50 towards Kv1.3 was determined by reported assays, were developed through exploration of four machine learning (ML) techniques. ML techniques explored included Random Forest, Support Vector Machine, Multilayer Perceptron, and K-Nearest Neighbors. Two QSAR classification approaches were developed. In the first approach, the compounds were classified into either moderate-active-to-very-active or inactive-to-moderate-active categories. Only the compounds predicted to be moderate-active-to-very-active in the first classification model were submitted to the second model, a classification model that predicted two more categories, very-active and not-very-active. The performances of the models were successfully evaluated by internal validation and external test set validation, with an overall predictability (Q) of 0.83 and 0.69 for the test set in the first and second approaches of the best models, respectively. The best models for the two approaches were employed for the virtual screening of KV1.3 inhibitors from ZINC natural products and approved drugs databases. A list of the most promising lead-like KV1.3 inhibitors was proposed, from which an approved drug and a natural product were experimentally evaluated with whole-cell voltage-clamp assays at 1000 nM.

AB - Voltage-gated ion channels are key molecular targets for autoimmune diseases such as multiple sclerosis, rheumatoid arthritis and psoriasis. In silico models, using 340 molecules whose IC50 towards Kv1.3 was determined by reported assays, were developed through exploration of four machine learning (ML) techniques. ML techniques explored included Random Forest, Support Vector Machine, Multilayer Perceptron, and K-Nearest Neighbors. Two QSAR classification approaches were developed. In the first approach, the compounds were classified into either moderate-active-to-very-active or inactive-to-moderate-active categories. Only the compounds predicted to be moderate-active-to-very-active in the first classification model were submitted to the second model, a classification model that predicted two more categories, very-active and not-very-active. The performances of the models were successfully evaluated by internal validation and external test set validation, with an overall predictability (Q) of 0.83 and 0.69 for the test set in the first and second approaches of the best models, respectively. The best models for the two approaches were employed for the virtual screening of KV1.3 inhibitors from ZINC natural products and approved drugs databases. A list of the most promising lead-like KV1.3 inhibitors was proposed, from which an approved drug and a natural product were experimentally evaluated with whole-cell voltage-clamp assays at 1000 nM.

KW - K1.3 inhibitors

KW - machine learning (ML)

KW - quantitative structure activity relationships (QSAR)

KW - random forest (RF)

KW - virtual screening

UR - http://www.scopus.com/inward/record.url?scp=85041542594&partnerID=8YFLogxK

U2 - 10.1002/slct.201702977

DO - 10.1002/slct.201702977

M3 - Article

VL - 3

SP - 1352

EP - 1364

JO - ChemistrySelect

JF - ChemistrySelect

SN - 2365-6549

IS - 5

ER -