A comparison of different electrostatic potentials on prediction accuracy in CoMFA and CoMSIA studies

Keng Chang Tsai, Yu Chen Chen, Nai Wan Hsiao, Chao Li Wang, Chih Lung Lin, Yu Ching Lee, Minyong Li, Binghe Wang

Research output: Contribution to journalArticle

33 Citations (Scopus)

Abstract

Computational chemistry is playing an increasingly important role in drug design and discovery, structural biology, and quantitative structure-activity relationship (QSAR) studies. For QSAR work, selecting an appropriate and accurate method to assign the electrostatic potentials of each atom in a molecule is a critical first step. So far several commonly used methods are available to assign charges. However, no systematic comparison of the effects of electrostatic potentials on QSAR quality has been made. In this study, twelve semi-empirical and empirical charge-assigning methods, AM1, AM1-BCC, CFF, Del-Re, Formal, Gasteiger, Gasteiger-Hückel, Hückel, MMFF, PRODRG, Pullman, and VC2003 charges, have been compared for their performances in CoMFA and CoMSIA modeling using several standard datasets. Some charge assignment models, such as Del-Re, PRODRG, and Pullman, are limited to specific atom and bond types, and, therefore, were excluded from this study. Among the remaining nine methods, the Gasteiger-Hückel charge, though commonly used, performed poorly in prediction accuracy. The AM1-BCC method was better than most charge-assigning methods based on prediction accuracy, though it was not successful in yielding overall higher cross-validation correlation coefficient (q2) values than others. The CFF charge model worked the best in prediction accuracy when q2 was used as the evaluation criterion. The results presented should help the selection of electrostatic potential models in CoMFA and CoMSIA studies.

Original languageEnglish
Pages (from-to)1544-1551
Number of pages8
JournalEuropean Journal of Medicinal Chemistry
Volume45
Issue number4
DOIs
Publication statusPublished - Apr 1 2010
Externally publishedYes

Fingerprint

Static Electricity
Electrostatics
Quantitative Structure-Activity Relationship
Computational chemistry
Atoms
Molecules
Drug Design
Drug Discovery
Pharmaceutical Preparations

Keywords

  • 3D-QSAR
  • AM1-BCC
  • CoMFA
  • CoMSIA
  • Electrostatic potentials

ASJC Scopus subject areas

  • Pharmacology
  • Drug Discovery
  • Organic Chemistry

Cite this

A comparison of different electrostatic potentials on prediction accuracy in CoMFA and CoMSIA studies. / Tsai, Keng Chang; Chen, Yu Chen; Hsiao, Nai Wan; Wang, Chao Li; Lin, Chih Lung; Lee, Yu Ching; Li, Minyong; Wang, Binghe.

In: European Journal of Medicinal Chemistry, Vol. 45, No. 4, 01.04.2010, p. 1544-1551.

Research output: Contribution to journalArticle

Tsai, Keng Chang ; Chen, Yu Chen ; Hsiao, Nai Wan ; Wang, Chao Li ; Lin, Chih Lung ; Lee, Yu Ching ; Li, Minyong ; Wang, Binghe. / A comparison of different electrostatic potentials on prediction accuracy in CoMFA and CoMSIA studies. In: European Journal of Medicinal Chemistry. 2010 ; Vol. 45, No. 4. pp. 1544-1551.
@article{1d47dabe049b4b73ad51afe7d7bc1e42,
title = "A comparison of different electrostatic potentials on prediction accuracy in CoMFA and CoMSIA studies",
abstract = "Computational chemistry is playing an increasingly important role in drug design and discovery, structural biology, and quantitative structure-activity relationship (QSAR) studies. For QSAR work, selecting an appropriate and accurate method to assign the electrostatic potentials of each atom in a molecule is a critical first step. So far several commonly used methods are available to assign charges. However, no systematic comparison of the effects of electrostatic potentials on QSAR quality has been made. In this study, twelve semi-empirical and empirical charge-assigning methods, AM1, AM1-BCC, CFF, Del-Re, Formal, Gasteiger, Gasteiger-H{\"u}ckel, H{\"u}ckel, MMFF, PRODRG, Pullman, and VC2003 charges, have been compared for their performances in CoMFA and CoMSIA modeling using several standard datasets. Some charge assignment models, such as Del-Re, PRODRG, and Pullman, are limited to specific atom and bond types, and, therefore, were excluded from this study. Among the remaining nine methods, the Gasteiger-H{\"u}ckel charge, though commonly used, performed poorly in prediction accuracy. The AM1-BCC method was better than most charge-assigning methods based on prediction accuracy, though it was not successful in yielding overall higher cross-validation correlation coefficient (q2) values than others. The CFF charge model worked the best in prediction accuracy when q2 was used as the evaluation criterion. The results presented should help the selection of electrostatic potential models in CoMFA and CoMSIA studies.",
keywords = "3D-QSAR, AM1-BCC, CoMFA, CoMSIA, Electrostatic potentials",
author = "Tsai, {Keng Chang} and Chen, {Yu Chen} and Hsiao, {Nai Wan} and Wang, {Chao Li} and Lin, {Chih Lung} and Lee, {Yu Ching} and Minyong Li and Binghe Wang",
year = "2010",
month = "4",
day = "1",
doi = "10.1016/j.ejmech.2009.12.063",
language = "English",
volume = "45",
pages = "1544--1551",
journal = "European Journal of Medicinal Chemistry",
issn = "0223-5234",
publisher = "Elsevier Masson SAS",
number = "4",

}

TY - JOUR

T1 - A comparison of different electrostatic potentials on prediction accuracy in CoMFA and CoMSIA studies

AU - Tsai, Keng Chang

AU - Chen, Yu Chen

AU - Hsiao, Nai Wan

AU - Wang, Chao Li

AU - Lin, Chih Lung

AU - Lee, Yu Ching

AU - Li, Minyong

AU - Wang, Binghe

PY - 2010/4/1

Y1 - 2010/4/1

N2 - Computational chemistry is playing an increasingly important role in drug design and discovery, structural biology, and quantitative structure-activity relationship (QSAR) studies. For QSAR work, selecting an appropriate and accurate method to assign the electrostatic potentials of each atom in a molecule is a critical first step. So far several commonly used methods are available to assign charges. However, no systematic comparison of the effects of electrostatic potentials on QSAR quality has been made. In this study, twelve semi-empirical and empirical charge-assigning methods, AM1, AM1-BCC, CFF, Del-Re, Formal, Gasteiger, Gasteiger-Hückel, Hückel, MMFF, PRODRG, Pullman, and VC2003 charges, have been compared for their performances in CoMFA and CoMSIA modeling using several standard datasets. Some charge assignment models, such as Del-Re, PRODRG, and Pullman, are limited to specific atom and bond types, and, therefore, were excluded from this study. Among the remaining nine methods, the Gasteiger-Hückel charge, though commonly used, performed poorly in prediction accuracy. The AM1-BCC method was better than most charge-assigning methods based on prediction accuracy, though it was not successful in yielding overall higher cross-validation correlation coefficient (q2) values than others. The CFF charge model worked the best in prediction accuracy when q2 was used as the evaluation criterion. The results presented should help the selection of electrostatic potential models in CoMFA and CoMSIA studies.

AB - Computational chemistry is playing an increasingly important role in drug design and discovery, structural biology, and quantitative structure-activity relationship (QSAR) studies. For QSAR work, selecting an appropriate and accurate method to assign the electrostatic potentials of each atom in a molecule is a critical first step. So far several commonly used methods are available to assign charges. However, no systematic comparison of the effects of electrostatic potentials on QSAR quality has been made. In this study, twelve semi-empirical and empirical charge-assigning methods, AM1, AM1-BCC, CFF, Del-Re, Formal, Gasteiger, Gasteiger-Hückel, Hückel, MMFF, PRODRG, Pullman, and VC2003 charges, have been compared for their performances in CoMFA and CoMSIA modeling using several standard datasets. Some charge assignment models, such as Del-Re, PRODRG, and Pullman, are limited to specific atom and bond types, and, therefore, were excluded from this study. Among the remaining nine methods, the Gasteiger-Hückel charge, though commonly used, performed poorly in prediction accuracy. The AM1-BCC method was better than most charge-assigning methods based on prediction accuracy, though it was not successful in yielding overall higher cross-validation correlation coefficient (q2) values than others. The CFF charge model worked the best in prediction accuracy when q2 was used as the evaluation criterion. The results presented should help the selection of electrostatic potential models in CoMFA and CoMSIA studies.

KW - 3D-QSAR

KW - AM1-BCC

KW - CoMFA

KW - CoMSIA

KW - Electrostatic potentials

UR - http://www.scopus.com/inward/record.url?scp=77349088235&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77349088235&partnerID=8YFLogxK

U2 - 10.1016/j.ejmech.2009.12.063

DO - 10.1016/j.ejmech.2009.12.063

M3 - Article

VL - 45

SP - 1544

EP - 1551

JO - European Journal of Medicinal Chemistry

JF - European Journal of Medicinal Chemistry

SN - 0223-5234

IS - 4

ER -