Knowing the 3D shape of the protein is important for trying to understand how they work and respond to drug therapies.
“Designing successful drugs is like solving a puzzle,” Ali Punjani, one of the algorithm developers, said in a news release.
“Without knowing the three-dimensional shape of a protein, it would be like trying to solve that puzzle with a blindfold on,” said Punjani, a University of Toronto PhD student.
Drugs attach to specific protein molecules and change their 3D shapes to affect how they work inside the body. The drugs are designed to be able to attach to only one type of protein molecule or proteins that come with diseases and are supposed to have no side effects.
“To design new drug therapies for diseases that haven’t been cured, we need to know the 3D structure of those protein molecules that are involved in each disease,” said Punjani in a University of Toronto video about the study.
The algorithms are designed to reconstruct 3D structures of protein molecules by using microscopic images. The already tiny protein molecules can only be seen with advanced techniques like electron cryomicroscopy (cryo-EM).
Cryo-EM takes tens of thousands of low-resolution images of frozen protein samples from different positions using high-power microscopes. The algorithm can then use those pictures to create a high-resolution 3D structure.
“Our approach solves some of the major problems in terms of speed and number of structures you can determine,” said professor David Fleet, chair of the computer and mathematical sciences department at the University of Toronto Scarborough and Punjani’s supervisor in the PhD program.
The algorithms were developed in part by one of Fleet’s former post-doctoral researchers, Marcus Brubaker, and could help develop new drugs because of its faster and more efficient method of creating correct protein structures.
“Existing techniques take several days or even weeks to generate a 3D structure on a cluster of computers,” said Brubaker. “Our approach can make it possible in minutes on a single computer.”
The user of the algorithm has to provide a guess of the molecule that is being studied for it to create the correct type of structure.
“We hope this will allow discoveries to happen at a ground-breaking pace in structural biology,” said Punjani. “The ultimate goal is that it will directly lead to new drug candidates for diseases and a much deeper understanding of how life works at the atomic level.”
The research was funded by the University of Toronto’s Innovations and Partnership’s Office through a Connaught Innovation Award, the University of Toronto’s Early Stage Technologies program, the Ontario Centers of Excellence and FedDev Ontario’s Investing in Commercialization Partnerships program with York University. It was published online in the Nature Methods journal.
[Want to stay more on top of MDO content? Subscribe to our weekly e-newsletter.]