AI-driven method developed to identify therapeutic targets

Artificial Intelligence Brain Social

Researchers have developed a method of identifying therapeutic targets for human diseases associated with protein phase separation (PPS) that they say could be used to discover targets for Alzheimer’s and other neurogenerative diseases.

The work is being presented as a “milestone” in the collaboration begun in 2021 between University of Cambridge academics and Insilico Medicine. Insilico has developed artificial intelligence (AI) platforms that utilize deep generative models and other modern machine-learning techniques to discover novel targets. Scientists say that exploring the process of PPS in cellular functions has been challenging.

"Even more difficult has been to clarify the exact nature of its association with human disease,” said Prof. Michele Vendruscolo, co-director of the Centre for Misfolding Diseases, Yusuf Hamied Department of Chemistry, at the University of Cambridge.

“By working with Insilico Medicine, we have developed a multiomic approach to systematically address this problem and identify a variety of possible targets for therapeutic intervention. We have thus provided a roadmap for researchers to navigate this complex terrain,” he said.

Their approach is explained in a paper published in the Proceedings of the National Academy of Sciences in which Vendruscolo is the lead author.

The dysregulation of the PPS process has been associated with numerous human diseases. It is postulated that PPS at the wrong place or time creates "clogs" or aggregates of molecules linked to neurodegenerative diseases; poorly formed cellular condensates could contribute to cancers and might help explain the aging process.

The researchers calculated the propensity of a protein to undergo spontaneous phase separation, which aims to aid in the identification of proteins prone to forming liquid-liquid phase-separated condensates. The researchers carried out a multiomic analysis to assess the likelihood of different diseases to benefit from PPS-based therapeutics.

The approach combined Insilico’s AI-driven target identification platform PandaOmics, which incorporates multiomic data for ranking genes based on their disease association, with the FuzDrop method introduced at Cambridge University of identifying proteins with a high probability of undergoing spontaneous PPS.

They quantified the relative impact of PPS in regulating various pathological processes associated with human disease, prioritized candidates with high PandaOmics and FuzDrop scores, and generated a list of possible therapeutic targets for human diseases linked with PPS. They then validated the differential phase separation behaviors of three predicted Alzheimer’s disease targets (MARCKS, CAMKK2, and p62) in two cell models of the disease.

The researchers say further studies are required to understand the molecular mechanisms that regulate condensate formation and their implications in the progression of the pathology.

“The study is intended to provide initial directions for targeting PPS-prone disease-associated proteins. With ongoing technical advancements in studying the PPS process, coupled with growing data about its roles in both cellular function and dysfunction, it is now possible to comprehend the causal relationship between PPS targets and diseases, said Frank Pun, PhD, head of Insilico Medicine Hong Kong and co-author of the paper.

“We anticipate facilitating the translation of this preclinical research into novel therapeutic interventions in the near future,” he added.

Page 1 of 17
Next Page