Format

Send to

Choose Destination
Cell Syst. 2015 Dec 23;1(6):383-95. doi: 10.1016/j.cels.2015.12.003. Epub 2015 Dec 23.

Prediction of Synergism from Chemical-Genetic Interactions by Machine Learning.

Author information

1
Wellcome Trust Centre for Cell Biology, School of Biological Sciences, University of Edinburgh, Edinburgh EH9 3JR, UK.
2
Wellcome Trust Centre for Cell Biology, School of Biological Sciences, University of Edinburgh, Edinburgh EH9 3JR, UK; Michael G. DeGroote Institute for Infectious Disease Research, McMaster University, Hamilton, ON L8N 3Z5, Canada.
3
Lunenfeld-Tanenbaum Research Institute, Mount Sinai Hospital, Toronto, ON M5G 1X5, Canada.
4
Michael G. DeGroote Institute for Infectious Disease Research, McMaster University, Hamilton, ON L8N 3Z5, Canada.
5
Wellcome Trust Centre for Cell Biology, School of Biological Sciences, University of Edinburgh, Edinburgh EH9 3JR, UK; Institute for Research in Immunology and Cancer, Department of Medicine, Université de Montréal, Montréal, QC H3C 3J7, Canada. Electronic address: md.tyers@umontreal.ca.

Abstract

The structure of genetic interaction networks predicts that, analogous to synthetic lethal interactions between non-essential genes, combinations of compounds with latent activities may exhibit potent synergism. To test this hypothesis, we generated a chemical-genetic matrix of 195 diverse yeast deletion strains treated with 4,915 compounds. This approach uncovered 1,221 genotype-specific inhibitors, which we termed cryptagens. Synergism between 8,128 structurally disparate cryptagen pairs was assessed experimentally and used to benchmark predictive algorithms. A model based on the chemical-genetic matrix and the genetic interaction network failed to accurately predict synergism. However, a combined random forest and Naive Bayesian learner that associated chemical structural features with genotype-specific growth inhibition had strong predictive power. This approach identified previously unknown compound combinations that exhibited species-selective toxicity toward human fungal pathogens. This work demonstrates that machine learning methods trained on unbiased chemical-genetic interaction data may be widely applicable for the discovery of synergistic combinations in different species.

KEYWORDS:

Bayesian analysis; antifungal; bipartite graph; chemical-genetic interaction; combination; genetic network; machine learning; random forest; synergism

Supplemental Content

Full text links

Icon for Elsevier Science Icon for PubMed Central
Loading ...
Support Center