Figuring out the 3D shapes of biological molecules is one of the most complex tasks of modern biology.
Numerous companies and research teams spend tons of money and countless hours to determine a molecular structure, but even so, the efforts are often unsuccessful.
Eismann and Raphael Townshend, two Stanford University Ph.D. students, under the careful leadership of Ron Dror, associate professor of computer science, developed an approach that puts an end to the existing problems by using computers to predict accurate structures.
Their approach is particularly successful even when learning from bits of known structures, making it usable to the variations of molecules whose structures are most challenging to determine experimentally.
Their work was sustained by two papers speaking of applications for RNA molecules and multi-protein complexes, which were made public in Science and Proteins.
The paper posted in Science is made in partnership with the Stanford laboratory of Rhiju Das, an associate professor of biochemistry.
“Structural biology, which is the study of the shapes of molecules, has this mantra that structure determines function,” Townshend stated.
The algorithm can precisely predict molecular structure, which, in return, can help scientists figure out how various molecules work, with numerous applications, ranging from fundamental biological research to informed drug design practices.
“Proteins are molecular machines that perform all sorts of functions. To execute their functions, proteins often bind to other proteins. If you know that a pair of proteins is implicated in a disease and you know how they interact in 3-D, you can try to target this interaction very specifically with a drug,” Eisman explained.
Townshend and Eismann are the Science paper’s co-lead authors, with Stanford postdoctoral scholar Andrew Watkins from Das lab, but also co-lead authors of the Proteins paper, in collaboration with Nathaniel Thomas, a former Stanford Ph.D.