Nonlinear Prediction of Quantitative Structure−Activity Relationships

Peter Tiño*
School of Computer Science, Birmingham University, Birmingham B15 2TT, U.K.
Ian T. Nabney
Neural Computing Research Group, Aston University, Birmingham B4 7ET, U.K.
Bruce S. Williams and Jens Lösel
Pfizer Global Research and Development, Sandwich, Kent CT13 9NJ, U.K.
Yi Sun
Faculty of Engineering and Information Sciences, University of Hertfordshire, Hatfield, AL10 9AB, U.K.
J. Chem. Inf. Comput. Sci., 2004, 44 (5), pp 1647–1653
DOI: 10.1021/ci034255i
Publication Date (Web): July 28, 2004
Copyright © 2004 American Chemical Society
*

 Corresponding author e-mail:  p.tino@cs.bham.ac.uk.

Abstract

Predicting the log of the partition coefficient P is a long-standing benchmark problem in Quantitative Structure−Activity Relationships (QSAR). In this paper we show that a relatively simple molecular representation (using 14 variables) can be combined with leading edge machine learning algorithms to predict logP on new compounds more accurately than existing benchmark algorithms which use complex molecular representations.

View: Full Text HTML | Hi-Res PDF

Tools

SciFinder Links

SciFinder subscribers: Click to sign in | Not a SciFinder subscriber? Learn more at www.cas.org

History

  • Published In Issue September 27, 2004
  • Received November 7, 2003

Recommend & Share

Related Content

Other ACS content by these authors: