qsar.md

Empirical Modeling of Quantitative Structure-Activity Relationships (QSAR)

Machine learning methods used: random forest; stochastic gradient boosting; association rule learning.

Journal Publications (Refereed)

• J. Ma, C. Tong, A. Liaw, R. Sheridan, J. Szumiloski, and V. Svetnik, 2009: Generating hypotheses about molecular structure-activity relationships (SARs) by solving an optimization problem. Statistical Analysis and Data Mining, 2: 161-174. 🔒

  • Published in the Special Issue on Chemogenomics, Part II.

• V. Svetnik, T. Wang, C. Tong, A. Liaw, R. P. Sheridan, and Q. Song, 2005: Boosting: an ensemble learning tool for compound classification and QSAR modeling. Journal of Chemical Information and Modeling, 45: 786-799. 🔒

• V. Svetnik, A. Liaw, C. Tong, J. C. Culberson, R. P. Sheridan, and B. P. Feuston, 2003: Random forest: a classification and regression tool for compound classification and QSAR modeling. Journal of Chemical Information and Computer Sciences, 43: 1947-1958. 🔒

  • This is by far the most frequently cited journal paper I have co-authored. According to the journal's twitter account (July 16, 2020 tweet), this was the third most cited paper published by this journal.

Conference Papers (Refereed)

• V. Svetnik, A. Liaw, C. Tong, and T. Wang, 2004: Application of Breiman's random forest to modeling structure-activity relationships of pharmaceutical molecules. Multiple Classifier Systems, Fifth International Workshop, MCS 2004, Proceedings, 9-11 June 2004, Cagliari, Italy; F. Roli, J. Kittler, and T. Windeatt (eds.). Lecture Notes in Computer Science, vol. 3077. Berlin: Springer, 334-343. 🔒

  • This is by far the most frequently cited conference paper I have co-authored.

Conference Papers (Unrefereed)

• V. Svetnik, A. Liaw, and C. Tong, 2004: Variable selection in random forest with application to quantitative structure-activity relationship. IIASS International School on Neural Networks "E. R. Caianiello" 7th Course: Ensemble Methods for Learning Machines, 22-28 September 2002, Vietri sul Mare, Salerno, Italy. (I am not certain of the accuracy of this citation.)

  • IIASS = International Institute for Advanced Scientific Studies.

• C. Tong, V. Svetnik, and A. Liaw, 2003: Quantitative structure-activity-relationship modeling using Leo Breiman's random forest. Joint Statistical Meetings 2003, 3-7 August 2003, San Francisco, California. American Statistical Association 2003 Proceedings, Section on Physical and Engineering Sciences, 4263-4266.

  • Runner-up for the JSM 2003 Oustanding Presentation Award, from the ASA Section on Physical and Engineering Sciences.

Conference Abstracts

• M. Torrent, C. Tong, A. Liaw, R. Nachbar, Y. Gao, R. Mosley, and C. Culberson, 2007: Molecular modeling-assisted attenuation of undesirable PXR activity. In Silico ADMET Conference: The Role of Protein-Structure Information in ADMET Prediction, 16-17 May 2007, London, U.K.

• A. Liaw, C.Tong, T.-C. Wang, and V. Svetnik, 2006: How to find drugs with trees: applications of ensemble methods in QSAR modeling. 2006 ENAR Spring Meeting, 26-29 March 2006, Tampa, FL.

  • ENAR = Eastern North American Region of the International Biometric Society.

• S. Ha, C. Tong, R. Perlow-Poehnelt, J. H. Lin, J. C. Culberson, R. P. Sheridan, and J. Hochman, 2005: QSAR models for predicting p-glycoprotein activity of antagonists for a GPCR target. 230th American Chemical Society National Meeting and Exposition, 28 August - 1 September 2005, Washington, D.C., MEDI 146A.22.

• V. Svetnik, T. Wang, C. Tong, and A. Liaw, 2005: Application of ensemble learning for modeling of quantitative structure-activity relations of pharmaceutical molecules. Joint Annual Meeting of the Interface Foundation of North America and the Classification Society of North America, 8-12 June 2005, Saint Louis, Missouri.

• C. Tong, V. Svetnik, A. Liaw, R. P. Sheridan, J. C. Culberson, B. P. Feuston, R. Evers, and D. Hartley, 2003: QSAR prediction of ADME parameters using a new machine learning tool--random forest. Predictive ADME, 17-18 November 2003, Boston, Massachusetts.

  • Media coverage: "Christopher Tong...discussed the application of the machine learning tool Random Forest for QSAR prediction of ADME parameters. This approach was found to be an accurate means of predicting blood-brain barrier permeability and P-gp transport of test compounds, while resisting overfitting and being robust to parameter tuning and noise." -- C. Watson, Drug Discovery Today: Biosilico, 2 (2): 55-56 (March 2004).

Notes

ADME = Absorption, Distribution, Metabolism, and Excretion.

ADMET = same, except add Toxicity.

(c) 2022-2024 by Christopher Tong