Researchers have utilized AI to understand how proteins behave within cells, offering insights that could aid in drug development.
A team of researchers at the University of Cambridge has created a detailed atlas of proteins to elucidate their behavior within human cells. This innovative tool has the potential to uncover the root causes of diseases linked to protein malfunctions, such as dementia and various cancers.
The researchers, whose work was published in Nature Communications, have identified new proteins inside cells that play crucial roles in a variety of bodily functions. The team concentrated on condensates, which are droplet-like structures within cells serving as meeting points for proteins to gather and organize themselves. These condensates are also pivotal sites where disease processes initiate.
The findings are openly available to scientists worldwide, enabling them to explore specific protein targets of interest and their associated condensate systems.
“Through this model, we have uncovered novel elements within membraneless cellular compartments and unearthed fundamental principles governing their functions,” stated Professor Tuomas Knowles, the lead researcher.
Protein Condensates
Cells are composed of meticulously organized molecules, with condensates serving as one mechanism for this organization. These microscopic hubs are essential components of the machinery that drives cellular functions.
“Up until now, we lacked a comprehensive map detailing which proteins congregate within specific condensates. Our work presents the first atlas in this regard,” explained Knowles.
Utilizing AI
Given the incomplete understanding of the rules guiding proteins within cells, the team embarked on creating this atlas to predict protein interactions within condensates.
“Our motivation for this research stemmed from the quest to comprehend the intricate nature of protein condensates beyond the existing scientific knowledge,” stated Dr. Kadi Liis Saar, the first author of the study and a postdoctoral fellow at the Centre for Misfolding Diseases.
The researchers leveraged extensive databases like StringDB and BioGRID, containing diverse cell-related data, in conjunction with detailed case studies on individual condensates.
The utilization of AI enables scientists to integrate and analyze vast and complex data sets that were previously challenging to compare. Whereas earlier studies focused on a limited number of proteins, this atlas offers a comprehensive view of cellular landscapes.
“Through this atlas, we can now predict the localization of every single protein within a cell, its interactions with other proteins, and its spatial distribution, facilitating new research opportunities and potential interventions in diseases linked to irregular condensate formation,” remarked Saar.
Protein Discoveries
The AI algorithm detected proteins within the cellular model that had not been previously identified. Confirming the presence of these proteins in a laboratory setting would validate the accuracy of the AI predictions.
“Our study revealed proteins within condensates that were previously unidentified in these locations. These proteins are involved in critical bodily functions such as fat distribution, actin formation in cells, and protein synthesis. Notably, these proteins were absent from the previous training dataset we utilized,” stated the researchers.
“These findings hold promise for advancing our understanding of the roles played by condensates in biological processes and the underlying physical mechanisms governing their formation.”
