Integrating Molecular Similarity and AlphaFold-Based Structural Alignment for Target Discovery in Trypanosoma cruzi

Background: Chagas disease, caused by the parasite Trypanosoma cruzi, remains a major neglected tropical disease, with millions of people living with the infection worldwide. Current treatments are effective in the acute stage of the disease, but are poorly tolerated and show reduced efficacy in chronic infections, highlighting an urgent need for novel therapeutic strategies. A key bottleneck in early-stage drug discovery is target identification, which is traditionally dependent on costly and low-throughput experimental methods. Computational approaches offer a cost-effective and fast alternative to traditional methods. Methods: In this study, we present an integrated in silico pipeline that combines ligand-based and structure-based computational approaches to prioritize potential molecular targets for bioactive compounds against T. cruzi. The ligand-based component performed similarity searches across curated bioactivity databases containing known ligand-protein associations, and the most similar candidates were then further evaluated using a structure-based approach through pairwise structural alignment against the T. cruzi proteome from AlphaFold. Results: The pipeline was validated using eight compounds with known targets, successfully recovering the correct target in six cases. Additionally, two compounds with anti-T. cruzi activity but unknown mechanisms of action were analyzed to hypothesize their potential targets. Conclusions: Overall, the pipeline demonstrated moderate success, with limitations arising from challenges in handling novel chemotypes and poorly annotated targets. Nevertheless, its modular nature allows for an easy adaptation to other neglected tropical diseases, providing a flexible and cost-effective framework for early-stage target prioritization.