A high-content AI-powered pipeline for malaria drug discovery
Poster presented as part of the Crick BioImage Analysis Symposium 2023.
Malaria-causing Plasmodium sporozoites are deemed the most vulnerable parasite stage affecting humans. Sporozoites rely on a specific type of motility, termed gliding, to migrate from the skin inoculation site and actively invade hepatocytes in the liver. Exploring such dynamic phenotypes in the context of a drug screen is not trivial. To address this challenge, we have devised an innovative drug screening pipeline that leverages live imaging and AI to profile in vitro sporozoite motile behaviors. We used inexpensive bright-field microscopy to produce movies of sporozoites in treated and non-treated conditions. The individual parasites were segmented and tracked using TrackMate, and then classified using a CNN. Finally, the track features were computed by GeNePy3D and were used to identify phenotypic clusters.
Permission has been given by authors to upload to Crick Figshare. Copyright remains with the original authors.