Using Machine Learning to Improve Antimalarial Drug Delivery With Nanoparticles

Abstract

Malaria, caused by the Plasmodium parasite transmitted through Anopheles mosquito bites, poses a significant global health challenge. The emergence of drug resistance in antimalarial treatments exacerbates this problem, particularly in endemic regions like the WHO African Region. Despite there being several antimalarial drugs, malaria cases persist, necessitating innovative solutions. This paper explores the potential of nanotechnology and machine learning to address drug resistance in malaria treatment. Drawing on a diverse range of literature, this research investigates optimal pairings of antimalarial drugs with nanoparticles to enhance drug delivery efficiency. Leveraging machine learning algorithms, predictive models are developed to forecast the effectiveness of nanoparticle-drug combinations. A Python-based machine learning program utilizing datasets from the University of Coruna and the ChEMBL database is employed to predict antimalarial activity in drug-nanoparticle pairings. The study reveals promising results, indicating the potential of nanoformulated primaquine as effective antimalarial agents. The logistic regression model created from the results of the machine learning model demonstrates room for improvement, particularly in addressing class imbalances and incorporating feature engineering techniques. However, as seen in past research, machine learning proves to be beneficial in drug discovery and can thereby assist in finding a solution to the drug resistance problem in antimalarial drugs. Enhanced machine learning models, coupled with experimental validation, has the possibility to hold promise in accelerating the discovery of potent antimalarial treatments in future studies, thereby mitigating the global burden of malaria and advancing towards sustainable malaria control strategies.