Estimating the domain of applicability for machine learning QSAR models: A study on aqueous solubility of drug discovery molecules

Timon Sebastian Schroeter, Anton Schwaighofer, Sebastian Mika, Antonius Ter Laak, Detlev Suelzle, Ursula Ganzer, Nikolaus Heinrich, Klaus Muller

Research output: Contribution to journalArticle

18 Citations (Scopus)

Abstract

We investigate the use of different Machine Learning methods to construct models for aqueous solubility. Models are based on about 4000 compounds, including an in-house set of 632 drug discovery molecules of Bayer Schering Pharma. For each method, we also consider an appropriate method to obtain error bars, in order to estimate the domain of applicability (DOA) for each model. Here, we investigate error bars from a Bayesian model (Gaussian Process (GP)), an ensemble based approach (Random Forest), and approaches based on the Mahalanobis distance to training data (for Support Vector Machine and Ridge Regression models). We evaluate all approaches in terms of their prediction accuracy (in cross-validation, and on an external validation set of 536 molecules) and in how far the individual error bars can faithfully represent the actual prediction error.

Original languageEnglish
Pages (from-to)651-664
Number of pages14
JournalJournal of Computer-Aided Molecular Design
Volume21
Issue number12
DOIs
Publication statusPublished - 2007 Dec 1
Externally publishedYes

    Fingerprint

Keywords

  • Aqueous
  • Bayesian modeling
  • Decision tree
  • Distance
  • Domain of applicability
  • Drug discovery
  • Ensemble
  • Error bar
  • Error estimation
  • Gaussian Process
  • Machine learning
  • Random forest
  • Ridge regression
  • Solubility
  • Support vector machine

ASJC Scopus subject areas

  • Molecular Medicine

Cite this