Multi-task learning for pKa prediction

Grigorios Skolidis, Katja Hansen, Guido Sanguinetti, Matthias Rupp

Research output: Contribution to journalArticlepeer-review

Abstract

Many compound properties depend directly on the dissociation constants of its acidic and basic groups. Significant effort has been invested in computational models to predict these constants. For linear regression models, compounds are often divided into chemically motivated classes, with a separate model for each class. However, sometimes too few measurements are available for a class to build a reasonable model, e.g., when investigating a new compound series. If data for related classes are available, we show that multi-task learning can be used to improve predictions by utilizing data from these other classes. We investigate performance of linear Gaussian process regression models (single task, pooling, and multi-task models) in the low sample size regime, using a published data set (n = 698, mostly monoprotic, in aqueous solution) divided beforehand into 15 classes. A multi-task regression model using the intrinsic model of co-regionalization and incomplete Cholesky decomposition performed best in 85 % of all experiments. The presented approach can be applied to estimate other molecular properties where few measurements are available.
Original languageEnglish
Pages (from-to)883-895
Number of pages13
JournalJournal of computer-Aided molecular design
Volume26
Issue number7
DOIs
Publication statusPublished - 2012

Keywords

  • pKa prediction
  • Multi-task learning
  • Quantitative structure–property relationships
  • Gaussian processes

Fingerprint

Dive into the research topics of 'Multi-task learning for pKa prediction'. Together they form a unique fingerprint.

Cite this