U. of Michigan professor: 'We hope this research can lead to better medicines'

Researchers at the University of Michigan have made a groundbreaking discovery that they say dramatically increases the process of building molecules for pharmaceuticals, agrichemicals and materials with the help of artificial intelligence. According to a release by the university, the team was led by Tim Cernak, an assistant professor of medical chemistry and chemistry. 

The findings were published in Science on Feb. 3 after years of researching combining chemical synthesis and data science.

“Making a chemical structure that has atoms in just the right place to give you efficacious and nontoxic medicines, for instance, is tricky,” Cernak said. “It requires a chemical synthesis strategy grounded in the chemical building blocks you can actually buy and then stitch together using chemical reactions.”

This time-consuming process has been significantly sped up, as the team was able to achieve synthesis of a complex alkaloid in only three steps compared to the previous seven to 26 steps. The goal of the research was to streamline the complex process of synthesizing molecules. 

The assistance of AI and the innovative approach by Cernak and his team managed to achieve the same result in just three steps, showing a major advancement in efficiency, the university said.

Cernak compared the construction of the molecules to a game of chess where strategic moves need to be orchestrated to reach the desired outcome. Inspiration was taken from the graph theory and researchers developed a logical framework to accelerate the process. 

The team leveraged the SYNTHIA retrosynthesis software, which provides a database of molecular pathways and formulas. Using their proprietary algorithm to analyze the data, the researchers identified high-impact steps along the synthesis pathway, as well as steps that contributed to progress but were ultimately inefficient. 

The data-driven approach allows them to optimize the synthesis process while reducing the number of steps needed to obtain the desired molecule. Through leveraging AI and computational methods, the goal of the team is to overcome limitations and expedite the discovery of novel therapeutic compounds.

According to the study, their combined efforts have pushed the boundaries of scientific knowledge, showcasing the immense potential of integrating AI and data science in the field of chemical synthesis. 

In conclusion, the Michigan researchers say their successful integration of artificial intelligence and chemical synthesis represents a significant milestone in the race to develop new medicines. 

“We hope this research can lead to better medicines,” Cernak said. “So far, we have been limited in the molecular structures we can quickly access with chemical synthesis.” 

The co-authors of this study include Yingfu Lin, senior research fellow in pharmacy, Rui (Sam) Zhang, a doctoral student in chemistry, and Di Wang, a doctoral student in pharmacy.