Deep Learning-driven research for drug discovery: Tackling Malaria
Problem
Discovering new antimalarial drug candidates is slow and expensive, and parasites constantly develop resistance to current drugs.
Why it matters
Malaria affects over 216 million people and kills over 445,000 patients annually.
What the AI did
Deep learning QSAR models predicted antiplasmodial activity and cytotoxicity of untested compounds; the best models virtually screened a large chemical database and top predictions were experimentally tested.
Results
Two compounds, LabMol-149 and LabMol-152, showed potent antiplasmodial activity (EC50 <500 nM) and low cytotoxicity in mammalian cells.
Impact
AI-guided screening yielded experimentally validated antimalarial hit compounds, though no drug was developed.
Record
| Venue | PLoS Computational Biology |
|---|---|
| Published | 2020 |
| Work type | article |
| Authors | Bruno Junior Neves, Rodolpho C. Braga, Vinícius M. Alves, Marília N. N. Lima, Gustavo Capatti Cassiano, Eugene N. Muratov, Fábio Trindade Maranhão Costa, Carolina Horta Andrade |
| Citations | 63 (OpenAlex, at import time) |
| References | 70 |
| Model or method | deep learning QSAR models (binary and continuous) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1371/journal.pcbi.1007025 |
| Open access | gold |
| Keywords | malaria, deep learning, QSAR, virtual screening, drug discovery, Plasmodium falciparum |
Importance rating
major — millions of lives, an industry or a field
Abstract states malaria affects over 216 million people and kills over 445,000 annually, and calls new antimalarials a major global health priority; AI screening yielded two validated hit compounds.