Artificial intelligence is becoming increasingly important in drug discovery. Advances in the use of Big Data, learning algorithms and powerful computers have now enabled researchers at the University of Zurich (UZH) to better understand a serious metabolic disease.
Cystinosis is a rare lyosomal storage disorder affecting around 1 in 100,000 to 200,000 newborns worldwide. Nephropathic (non-inflammatory) cystinosis, the most common and severe form of the disease, manifests with kidney disease symptoms during the first months of life, often leading to kidney failure before the age of 10. “Children with cystinosis suffer from a devastating, multisystemic disease, and there are currently no available curative treatments,” says Olivier Devuyst, head of the Mechanisms of Inherited Kidney Disorders (MIKADO) group and co-director of the ITINERARE University Research Priority Program at UZH.
The UZH researchers worked with Insilico Medicine, a company that uses AI for drug discovery, to uncover the underlying cellular mechanism behind kidney disease in cystinosis. Leveraging model systems and Insilico’s PandaOmics platform, they identified the disease-causing pathways and prioritized therapeutic targets within cystinosis cells. Their findings revealed a causal association between the regulation of a protein called mTORC1 and the disease. Alessandro Luciani, one of the research group leaders, explains: “Our research showed that cystine storage stimulates the activation of the mTORC1 protein, leading to the impairment of kidney tubular cell differentiation and function.”