Stratifying antimalarial compounds with similar mode of action using machine learning on chemo-transcriptomic profiles

Abstract

Malaria is a terrible disease caused by a protozoan parasite within the Plasmodium genus, claiming the lives of hundreds of thousands of people yearly, the majority of whom are children under the age of five. Of the five species of Plasmodium causing malaria in humans, P. falciparum is responsible for most of the death toll. An increase in malaria cases was detected between the years 2016 to 2017 according to the World Malaria Report of 2017, despite control efforts. The rapid development of resistance within P. falciparum against antimalarials has led to the use of artemisinin combinational therapy as the current gold standard for malaria treatment. Yet decreased parasite clearance demonstrates that using combination therapy is insufficient in maintaining current antimalarials’ effectiveness against these resistant parasites. Hence, novel compounds with a mode of action (MoA) different than current antimalarials are required. Though phenotypic screening has delivered thousands of promising hit compounds, hit-to-lead optimisation is still one of the rate-limiting steps in pre-clinical antimalarial drug development. While knowing the exact target or MoA is not required to progress a compound in a medicinal chemistry program, identifying the MoA early can accelerate hit prioritization, hit-to-lead optimisation and preclinical combination studies in malaria research. In this study, we assessed machine learning (ML) approaches for their ability to stratify antimalarials based on transcriptional responses associated with the treatments. From our results, we conclude that it is possible to identify biomarkers from the transcriptional responses that define the MoA of compounds. Moreover, only a limited set of 50 genes was required to build a ML model that can stratify compounds with similar MoA with a classification accuracy of 76.6 ± 6.4%. These biomarkers will help stratify new compounds with similar MoA to those already defined with our strategy. Additionally, the biomarkers can also be used to monitor if the MoA of a compound has changed during hit-to-lead optimisation. This work will contribute to accelerating antimalarial drug discovery during the hit-to-lead optimisation phase and help the identification of compounds with novel MoA.