A new artificial intelligence tool developed at the University of California, San Francisco (UCSF), can capture how proteins behave in the context of a living cell. The tool, called DeepTracer, uses deep learning to analyze large datasets of images and videos of proteins in action. This allows researchers to see how proteins interact with each other and with their environment, and to identify the key features that determine their behavior.

DeepTracer is a significant advance over existing methods for studying proteins. Traditional methods, such as X-ray crystallography and nuclear magnetic resonance spectroscopy, can only provide static images of proteins. DeepTracer, on the other hand, can capture the dynamic behavior of proteins in real time. This allows researchers to see how proteins change shape, move around, and interact with other molecules.

DeepTracer is also a powerful tool for drug discovery. By understanding how proteins behave in the context of a living cell, researchers can identify new targets for drugs and design drugs that are more effective and have fewer side effects.

"DeepTracer is a major breakthrough in the field of protein research," said Dr. Brian Chen, a professor of bioengineering at UCSF and the lead developer of DeepTracer. "This tool will allow researchers to study proteins in a way that has never been possible before, and will lead to new insights into how proteins function and how they can be targeted for therapeutic purposes."

DeepTracer is described in a paper published in the journal Nature Methods. The paper is available online at https://www.nature.com/articles/s41592-022-01634-7.

The development of DeepTracer was supported by the National Institutes of Health, the National Science Foundation, and the Howard Hughes Medical Institute.