Title: In-silico study of computational modelling and GLP-1 receptor inverse agonist compounds on a cancer cell line inhibitory bioassay dataset

Authors: Harleen Kaur; Naved Ahmad; Ritu Chauhan

Addresses: Department of Computer Science, Hamdard University, New Delhi 110 062, India ' GD Research Centre Private Limited, Hyderabad 500081, India ' Amity Institute of Biotechnology, Amity University, Noida, U.P. 201313, India

Abstract: The introduction of technologies such as high-throughput screening (HTS) along with combinatorial chemistry has resulted in vast amounts of data generated from various experimental results. The existing automation technology such as machine learning tends to overcome the flaws of traditional mechanisms; hence, traditional analysis techniques tend to be inefficient, time consuming and costly when dealing with the complexity of large data. The objective of target-based modelling is prediction of the activity and relationships among different compounds from a large database with unknown activity and thus reducing the cost and time for discovery and development of a new drug. To deal with current objectives, we have discussed the comparative study for different machine learning algorithms such as naïve Bayes, random forest, sequential minimal optimisation (SMO) and J48 for generating predictive models. The approach was extended and evaluated to measure the accuracy of targeted models with statistical techniques to increase accuracy.

Keywords: predictive modelling; GLP-1 receptors; machine learning; docking; drug design; bioinformatics; inverse agonist compounds; cancer datasets; naive Bayes; random forest; sequential minimal optimisation; SMO; J48.

DOI: 10.1504/IJCBDD.2015.072116

International Journal of Computational Biology and Drug Design, 2015 Vol.8 No.3, pp.293 - 309

Received: 30 Aug 2014
Accepted: 04 Feb 2015
Published online: 30 Sep 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article