AI to decode the language of cancer and Alzheimer's disease
2021
University of Cambridge, Cambridge, United Kingdom
Intracellular phase separation of proteins into biomolecular condensates is increasingly recognized as a process with a key role in cellular compartmentalization and regulation. And dysfunction is seen as a trigger for cancer and neurodegenerative diseases such as Alzheimer's disease. To understand how protein sequence determines phase behaviour and to develop an algorithm to predict LLPS-prone sequences(liquid-liquid phase separation), datasets of proteins with different LLPS propensities were created. The DeePhase model showed high performance in both distinguishing LLPS-prone proteins from structured proteins and identifying them within the human proteome. Overall, the results shed light on the physicochemical factors that modulate protein condensation and provide a molecular principles-based platform for predicting protein phase behavior.
Learning the molecular grammar of protein condensates from sequence determinants and embeddings
Tuomas P. J. Knowles
Added on: 04-19-2021
[1] https://www.pnas.org/content/118/15/e2019053118[2] https://www.bionity.com/en/news/1170578/artificial-intelligence-could-crack-the-language-of-cancer-and-alzheimer-s.html