AI that can design new drug molecules from scratch developed by researchers

AI will dramatically accelerate the design of new drugs, claims University of North Carolina scientists

Researchers at the University of North Carolina claim to have created an artificial intelligence system that can teach itself to design new drug molecules from scratch.

Claiming to "dramatically accelerate the design of new drug candidates", the machine is called Reinforcement Learning for Structural Evolution, or 'ReLeaSE', for short.

It's basically an algorithm and computer program that comprises two neural networks that, the University says, can be best thought of as a teacher and a student.

"The teacher knows the syntax and linguistic rules behind the vocabulary of chemical structures for about 1.7 million known biologically active molecules," the university explained.

We want to give scientists a grocery store and a personal chef who can create any dish they want

"By working with the teacher, the student learns over time and becomes better at proposing molecules that are likely to be useful as new medicines."

The scientists behind the creation are the University's Eshelman School of Pharmacy professors Alexander Tropsha, Olexandr Isayev, and Mariya Popova.

Since developing it, they have applied for a patent for the technology, and published a proof-of-concept study in the journal Science Advances.

"If we compare this process to learning a language, then after the student learns the molecular alphabet and the rules of the language, they can create new 'words,' or molecules," said Tropsha.

"If the new molecule is realistic and has the desired effect, the teacher approves. If not, the teacher disapproves, forcing the student to avoid bad molecules and create good ones."

The ability of the algorithm to design new and, therefore, immediately patentable, chemical entities with specific biological activities and optimal safety profiles should be highly attractive

ReLeaSE, they suggested, will shake-up virtual screening, a technology that enables scientists to evaluate existing large chemical libraries. The problem with virtual screening is that the method only works for known chemicals. However, ReLeASE has the unique ability to create and evaluate new molecules, the researchers claimed.

"A scientist using virtual screening is like a customer ordering in a restaurant. What can be ordered is usually limited by the menu," said Isayev. "We want to give scientists a grocery store and a personal chef who can create any dish they want."

The team has therefore been able to use ReLeaSE to generate molecules with properties that they specified, such as desired bioactivity and safety profiles.

They used this method to design molecules with customised physical properties, such as melting point and solubility in water, and to design new compounds with inhibitory activity against an enzyme that is associated with leukemia.

"The ability of the algorithm to design new and, therefore, immediately patentable, chemical entities with specific biological activities and optimal safety profiles should be highly attractive to an industry that is constantly searching for new approaches to shorten the time it takes to bring a new drug candidate to clinical trials," Tropsha concluded.