Shared Consensus Machine Learning Models for Predicting Blood Stage Malaria Inhibition
Andreas Verras, Chris L. Waller, Peter Gedeck et al. · Journal of Chemical Information and Modeling · 2017
The development of new antimalarial therapies is essential, and lowering the barrier of entry for the screening and discovery of new lead compound classes can spur drug development at organizations that may not have large compound screening libraries or resources to conduct high-throughput screens. Machine learning models have been long established to be more robust and have a larger domain of applicability with larger training sets. Screens over multiple data sets to find compounds with potential malaria blood stage inhibitory activity have been used to generate multiple Bayesian models. Here we describe a method by which Bayesian quantitative structure-activity relationship models, which contain information on thousands to millions of proprietary compounds, can be shared between collaborators at both for-profit and not-for-profit institutions. This model-sharing paradigm allows for the development of consensus models that have increased predictive power over any single model and yet does not reveal the identity of any compounds in the training sets.
Cited by 25 works. doi:10.1021/acs.jcim.6b00572
Authors
- Andreas Verras
- Chris L. Waller
- Peter Gedeck
- Darren V. S. Green
- Thierry Kogej
- Anandkumar V. Raichurkar
- Manoranjan Panda
- Anang A. Shelat
- Julie A. Clark
- Rodney Kiplin Guy
- George Papadatos
- Jeremy Nicholas Burrows
References
- Extended-Connectivity Fingerprints
- The ChEMBL bioactivity database: an update
- Atomic physicochemical parameters for three-dimensional-structure-directed quantitative structure-activity relationships. 2. Modeling dispersive and hydrophobic interactions
- Thousands of chemical starting points for antimalarial lead identification
- Simple and Inexpensive Fluorescence-Based Technique for High-Throughput Antimalarial Drug Screening
- Evaluating Virtual Screening Methods: Good and Bad Metrics for the “Early Recognition” Problem
- Chemical genetics of Plasmodium falciparum
- Similarity to Molecules in the Training Set Is a Good Discriminator for Prediction Accuracy in QSAR
- What do we know and when do we know it?
- The importance of the domain of applicability in QSAR modeling