Machine learning speeds up new drug efficiency tests.

Researchers at Massachusetts Institute of Technology (Broad Institute) and Harvard University have developed a machine learning-based technique that speeds up calculations of drug molecules' binding affinity to proteins. This indicator allows to check whether the new drug will work.

The approach yields precise calculations in a fraction of the time compared to previous methods.

“Our method is orders of magnitude faster than before, meaning we can have drug discovery that is both efficient and reliable,” says Bin Zhang, a co-author of the research.

The affinity between a drug molecule and a target protein is measured by a quantity called the binding free energy — the smaller the number, the stickier the bind. Calculating the binding free energy of a drug candidate provides an indicator of a drug’s potential effectiveness.

One method for computing binding free energy calculates the quantity exactly, eating up significant time and computer resources. The second method is less computationally expensive, but it yields only an approximation of the binding free energy.

DeepBAR computes binding free energy exactly, but it requires just a fraction of the calculations demanded by previous methods. The new technique combines traditional chemistry calculations with recent advances in machine learning.

The “BAR” in DeepBAR stands for “Bennett acceptance ratio,” a decades-old algorithm used in exact calculations of binding free energy. Using the Bennet acceptance ratio typically requires a knowledge of two “endpoint” states (e.g., a drug molecule bound to a protein and a drug molecule completely dissociated from a protein), plus knowledge of many intermediate states (e.g., varying levels of partial binding), all of which bog down calculation speed.

DeepBAR slashes those in-between states by deploying the Bennett acceptance ratio in machine-learning frameworks called deep generative models. “These models create a reference state for each endpoint, the bound state and the unbound state,” says Zhang.

In using deep generative models, the researchers were borrowing from the field of computer vision which people use to do computer image synthesis. It allows treating each molecular structure as an image, which the model can learn.

The DeepBAR development also raised some challenges. Since these models were originally developed for 2D images, adapting those methods to 3D structures was the biggest technical challenge.

In tests using small protein-like molecules, DeepBAR calculated binding free energy nearly 50 times faster than previous methods. The researchers add that, in addition to drug screening, DeepBAR could aid protein design and engineering, since the method could be used to model interactions between multiple proteins. In the future, the researchers plan to improve DeepBAR’s ability to run calculations for large proteins.

Previously, a team from the Massachusetts Institute of Technology and the Institute of Data, Systems and Society has developed a machine learning approach to identify drugs already on the market that could potentially be repurposed to combat Covid-19. In addition, employees of MIT, Cambridge and Harvard universities proposed using artificial intelligence designed to analyze virus mutations for text recognition.