Google's DeepMind "solves" Protein Folding

[TwinIon]

The full article in Nature is worth reading, but here are some highlights.

 

There is a yearly protein structure prediction challenge called Critical Assessment of Structure Prediction (CASP) that plays out over several months. About 100 target proteins or portions of them are released on a regular basis, and teams have several weeks to submit their predictions for how the proteins are structured. A separate team uses methods such as X-ray crystallography and cryo-electron microscopy (cryo-EM) to determine the actual structure and compare those against the predictions. A score of ~90 is considered to have effectively predicted the structure, and Google's AlphaFold 2 algorithm just about hit that mark this year. In 2016, the average score of the field competing was ~40.

 

Quote

 

“This is a big deal,” says John Moult, a computational biologist at the University of Maryland in College Park, who co-founded CASP in 1994 to improve computational methods for accurately predicting protein structures. “In some sense the problem is solved.”

...

“It’s a game changer,” says Andrei Lupas, an evolutionary biologist at the Max Planck Institute for Developmental Biology in Tübingen, Germany, who assessed the performance of different teams in CASP. AlphaFold has already helped him find the structure of a protein that has vexed his lab for a decade, and he expects it will alter how he works and the questions he tackles. “This will change medicine. It will change research. It will change bioengineering. It will change everything,” Lupas adds.

 

 

Interestingly the DeepMind team first tried to basically apply their tech to what other teams at the time had been trying. They tried to predict the distance between pairs of amino acids and then developed a consensus model of what the structure should look like. That approach hit a wall, so they changed their approach to just try and predict the whole final structure from the get go. It's a far more difficult task, but they've had far better results.

 

It seems that even with our most advanced imaging techniques, we often still need a prediction of the structure of a protein in order to verify the experimental data.

 

Quote

An AlphaFold prediction helped to determine the structure of a bacterial protein that Lupas’s lab has been trying to crack for years. Lupas’s team had previously collected raw X-ray diffraction data, but transforming these Rorschach-like patterns into a structure requires some information about the shape of the protein. Tricks for getting this information, as well as other prediction tools, had failed. “The model from group 427 gave us our structure in half an hour, after we had spent a decade trying everything,” Lupas says.

 

The problem isn't really "solved," but with sufficiently accurate predictions now possible from a computer model, scientists can now move on to a new realm of molecular biology.

 

Quote

 

AlphaFold is unlikely to shutter labs, such as Brohawn’s, that use experimental methods to solve protein structures. But it could mean that lower-quality and easier-to-collect experimental data would be all that’s needed to get a good structure. Some applications, such as the evolutionary analysis of proteins, are set to flourish because the tsunami of available genomic data might now be reliably translated into structures. “This is going to empower a new generation of molecular biologists to ask more advanced questions,” says Lupas. “It’s going to require more thinking and less pipetting.”

“This is a problem that I was beginning to think would not get solved in my lifetime,” says Janet Thornton, a structural biologist at the European Molecular Biology Laboratory-European Bioinformatics Institute in Hinxton, UK, and a past CASP assessor. She hopes the approach could help to illuminate the function of the thousands of unsolved proteins in the human genome, and make sense of disease-causing gene variations that differ between people.

 

 

It's all very exciting, and great to see this kind of tech being used for something with such huge real world implications.

[Keyser_Soze]