Using Machine Learning to Predict Synergistic Antimalarial Compound Combinations With Novel Structures
Problem
Malaria resistance is developing rapidly, and exhaustively screening large drug spaces for synergistic antimalarial combinations experimentally is infeasible.
Why it matters
Malaria is lethal and novel treatment modalities are urgently needed.
What the AI did
CoSynE, a machine learning approach, predicted novel synergistic drug interactions from prior combination screening data and compound structures, applied to 1,540 antimalarial drug combinations.
Results
Cross-validation: synergistic predictions enriched 2.74× vs random selection; dataset of 1,540 combinations, 22.2% synergistic
Impact
Evidence supports ML enrichment of synergy prediction in cross-validation; experimental validation not shown in abstract.
Record
| Venue | Frontiers in Pharmacology |
|---|---|
| Published | 2018 |
| Work type | article |
| Authors | Daniel J. Mason, Richard T. Eastman, Richard P. I. Lewis, Ian P. Stott, Rajarshi Guha, Andreas Bender |
| Citations | 42 (OpenAlex, at import time) |
| References | 68 |
| Model or method | CoSynE (Combination Synergy Estimation) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3389/fphar.2018.01096 |
| Open access | gold |
| Keywords | machine learning, drug synergy, antimalarial, CoSynE, cross validation |
Importance rating
major — millions of lives, an industry or a field
Abstract states this is the most lethal Plasmodium species causing serious malaria, with rapid resistance development and urgent need for novel treatment modalities; ML enriches synergy discovery.