Random Forest Model Predictions Afford Dual-Stage Antimalarial Agents.

Area of Science:

• Medicinal Chemistry
• Parasitology
• Computational Biology

Background:

• Malaria remains a significant global health burden, necessitating new treatments.
• Existing antimalarials face challenges, including drug resistance and limited efficacy against all parasite stages.
• Dual-stage inhibitors targeting both liver and blood stages are crucial for comprehensive malaria control.

Purpose of the Study:

• To leverage machine learning for the discovery of novel small-molecule antimalarials.
• To identify compounds with efficacy against both liver and blood stages of the malaria parasite.
• To find drug candidates with low cytotoxicity to human liver cells.

Main Methods:

• Utilized a random forest modeling approach to predict antimalarial activity.
• Screened a commercial diversity library using predictive models.
• Validated predicted compounds through in vitro testing against Plasmodium parasites.
• Assessed cytotoxicity against human liver cell lines.

Main Results:

• Identified 18 validated small-molecule antimalarials with dual-stage efficacy.
• Compounds demonstrated in vitro activity against liver-stage Plasmodium berghei.
• A subset of compounds also showed efficacy against blood-stage Plasmodium falciparum.
• Validated compounds exhibited low cytotoxicity to human liver cells.

Conclusions:

• Machine learning effectively identified novel dual-stage antimalarial drug candidates.
• The validated compounds serve as promising starting points for further antimalarial drug discovery.
• These novel agents hold potential for developing a new generation of malaria therapies.