Potency-directed similarity searching using support vector machines
© Heikamp et al; licensee BioMed Central Ltd. 2012
Published: 1 May 2012
Searching for active compounds in screening databases is one of the main tasks in chemoinformatics [1, 2]. For this purpose, different approaches have been developed, with an increasing interest in machine learning and data mining methods . Among these, support vector machine (SVM) learning has proven to be a powerful search technique in many instances . Several applications of SVMs have been reported that focus on compound ranking in similarity searching [4–6]. However, similarity search and machine learning methods that are commonly utilized for virtual screening generally do not take compound potency information into account. Regardless of the applied methods, one typically attempts to distinguish “active” from “inactive” compounds. With the exception of QSAR models adapted for compound screening, no approaches have thus far been introduced that incorporate potency information as a parameter and direct search calculations toward the recognition of potent hits. Here, an SVM approach for potency-directed virtual screening is introduced . A newly designed structure-activity kernel and an SVM linear combination model take potency information of reference molecules into account. Applied to high-throughput screening data sets, this potency-directed SVM approach met or exceeded the recall performance of standard SVM ranking and led to a notable enrichment of highly potent hits in database selection sets.
- Stumpfe D, Bajorath J: Applied virtual screening: strategies, recommendations, and caveats. Methods and Principles in Medicinal Chemistry. Virtual Screening. Principles, Challenges, and Practical Guidelines. Edited by: Sotriffer C. 2011, Weinheim: Wiley-VCH, 73-103.Google Scholar
- Eckert H, Bajorath J: Molecular similarity analysis in virtual screening: foundations, limitations and novel approaches. Drug Discov Today. 2007, 12: 225-233. 10.1016/j.drudis.2007.01.011.View ArticleGoogle Scholar
- Geppert H, Vogt M, Bajorath J: Current trends in ligand-based virtual screening: molecular representations, data mining methods, new application areas, and performance evaluation. J Chem Inf Model. 2010, 50: 205-216. 10.1021/ci900419k.View ArticleGoogle Scholar
- Jorissen RN, Gilson MK: Virtual screening of molecular databases using a support vector machine. J Chem Inf Model. 2005, 45: 549-561. 10.1021/ci049641u.View ArticleGoogle Scholar
- Wassermann AM, Geppert H, Bajorath J: Searching for target-selective compounds using different combinations of multiclass support vector machine ranking methods, kernel functions, and fingerprint descriptors. J Chem Inf Model. 2009, 49: 582-592. 10.1021/ci800441c.View ArticleGoogle Scholar
- Agarwal S, Dugar D, Sengupta S: Ranking chemical structures for drug discovery: a new machine learning approach. J Chem Inf Model. 2010, 50: 716-731. 10.1021/ci9003865.View ArticleGoogle Scholar
- Wassermann AM, Heikamp K, Bajorath J: Potency-directed similarity searching using support vector machines. Chem Biol Drug Des. 2011, 77: 30-38. 10.1111/j.1747-0285.2010.01059.x.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.