Skip to main content
Kent Academic Repository

Evaluation of QSAR and ligand enzyme docking for the identification of ABCB1 substrates

Osho, V., Ojo, O., Sharifi, M., Ghafourian, T. (2013) Evaluation of QSAR and ligand enzyme docking for the identification of ABCB1 substrates. In: UK QSAR & Chemoinformatics Group Meeting, April 2013, Unilever, Bedford. (KAR id:42820)

PDF (Evaluation of QSAR and ligand enzyme docking for the identification of ABCB1 substrates) Presentation
Language: English
Download this file
(PDF/1MB)
[thumbnail of Evaluation of QSAR and ligand enzyme docking for the identification of ABCB1 substrates]
Preview
Request a format suitable for use with assistive technology e.g. a screenreader

Abstract

P-glycoprotein (P-gp) is an efflux pump that belongs to ATP-binding cassette (ABC) transporter family embedded in the membrane bilayer. P-gp is a polyspecific protein that has demonstrated its function as a transporter of hydrophobic drugs as well as transporting lipids, steroids and metabolic products. Its role in multidrug resistance (MDR) and pharmacokinetic profile of clinically important drug molecules has been widely recognised. In this study, QSAR and enzyme-ligand docking methods were explored in order to classify substrates and non-substrates of P-glycoprotein. A set of 123 compounds designated as substrates (54) or non-substrates (69) by Matsson et al., 2009 was used for the investigation. For QSAR studies, molecular descriptors were calculated using ACD labs/LogD Suite and MOE (CCG Inc.). P-glycoprotein structures available in the Protein data bank were used for docking studies and determination of binding scores using MOE software. Binding sites were defined using co-crystallised ligand structures. Three classification algorithms which included classification and regression trees, boosted trees and support vector machine were examined. Models were developed using a training set of 98 compounds and were validated using the remaining compounds as the external test set. A model generated using BT was identified as the best of three models, with a prediction accuracy of 88%, Mathews correlation coefficient of 0.77 and Youden’s J index of 0.80 for the test set. Inclusion of various docking scores for different binding sites improved the models only marginally.

Item Type: Conference or workshop item (Poster)
Subjects: Q Science > Q Science (General)
Q Science > QD Chemistry
R Medicine > R Medicine (General)
R Medicine > RM Therapeutics. Pharmacology
Divisions: Divisions > Division of Natural Sciences > Medway School of Pharmacy
Depositing User: Taravat Ghafourian
Date Deposited: 03 Sep 2014 17:13 UTC
Last Modified: 29 Dec 2022 21:02 UTC
Resource URI: https://kar.kent.ac.uk/id/eprint/42820 (The current URI for this page, for reference purposes)

University of Kent Author Information

Ghafourian, T..

Creator's ORCID:
CReDIT Contributor Roles:
  • Depositors only (login required):

Total unique views for this document in KAR since July 2020. For more details click on the image.