Precision through Networks: AI-driven Targeting of Protein-Protein Interactions in Lysosomal Storage Disorders

Understanding protein-protein interactions is crucial for deciphering cellular processes and diseases. Perturbation data offers insights into phenotypic alterations resulting from modifications to protein interactions, revealing functional relationships between proteins and pathways. These interactions shape cellular phenotypes and influence disease susceptibility or progression. Most protein-protein interactions are non-additive, adding to the complexity of phenotypic variations. Integrating diverse data modalities, such as gene expression profiles, genomic and protein sequences, and biological knowledge graphs, our proposed multi-modal artificial intelligence (AI) model aims to predict cellular responses to perturbation.

Our approach utilizes advanced machine learning algorithms to analyze and integrate high-dimensional datasets, enabling the identification of intricate patterns and interactions that traditional methods might overlook. By uncovering complex protein-protein interactions, our approach aims to identify novel therapeutic opportunities, particularly in rare lysosomal storage diseases affecting 1/5000 live births. These diseases often result from defects in lysosomal enzymes, leading to the accumulation of undigested substrates and subsequent cellular dysfunction.

Current therapies face significant challenges due to the systemic nature of these conditions, often requiring lifelong treatment. Our multimodal AI framework offers a promising solution by providing a comprehensive understanding of disease mechanisms at the molecular level. This can lead to the development of more effective, targeted therapies that address the root causes of lysosomal storage disorders rather than just managing symptoms. This project addresses the lack of effective therapeutic strategies for such diseases by revealing novel opportunities for intervention advancing targeted therapeutics.