UGA team awarded NIH funding to use AI to investigate molecular interactions

Molecular scale breakthroughs in human health – from COVID-19 vaccines to cancer therapeutics – require scientists to understand how molecules interact with each other, akin to figuring out how puzzle pieces fit together. To do this, scientists use a special microscope called an atomic force microscope (AFM) that can see and manipulate molecules. For the virus that causes COVID-19, for example, this microscope can gently poke and measure how strong the virus sticks to the body's cells. Scientists can use this to understand how the virus attaches to our cells and causes infection.

When these molecules interact, they exert forces on each other, similar to pulling two magnets apart. But with molecules, these forces are super small and very difficult to measure. Investigation and manipulation of single molecules requires understanding of the mechanical properties and force-induced structural rearrangements of sugars, DNA, and proteins. Atomic force microscopy-single-molecule force spectroscopy (AFM-SMFS) allows force probing of the surfaces of living cells at molecular resolution and provides information about cell surface receptors. 

Numerous algorithms have been developed to analyze the massive number of AFM-SMFS curves to avoid human bias and errors and to save time. 

In a blend of technology and research capacity, scientists have developed smart algorithms that analyze the data from the microscope and figure out how strong the forces are between the molecules. This helps scientists understand how these molecules work together and how they might behave in different situations.

Now, an interdisciplinary University of Georgia research team has been awarded by $1.2 million from the National Institute of General Medical Sciences of the National Institutes of Health to use AI algorithms to make an atomic force microscopy.

The NIH-funded project, led by Wenxuan Zhong, professor in the Franklin College of Arts and Science department of statistics, builds on a paper published in 2021 as the cover story in the Journal of Physical Chemistry B. describing development of a network-based automatic clustering algorithm (NASA), to decode the details of specific molecules.

"Using AI in this way has big implications. For example, it can help us learn about how proteins in our bodies fold and unfold, which is crucial for their function. It can also give us insights into how drugs interact with proteins and how cells communicate," said Ping Ma, Distinguished Research Professor of statistics and co-P.I. on the project. "AI helps scientists unlock the secrets of these tiny interactions, which has a ripple effect on our understanding of biology and medicine." 

"Atomic force microscopy-single-molecule force spectroscopy (AFM-SMFS) is a powerful methodology to probe intermolecular and intramolecular interactions in biological systems because of it work so well under conditions the body, its simple and rapid sample preparation, and the combined functionality with high-resolution imaging," Zhong said.

Since a huge number of AFM-SMFS force−distance curves are collected to avoid human bias and errors and to save time, numerous algorithms have been developed to analyze the AFM-SMFS curves. Nevertheless, there is still a need to develop new algorithms for the analysis of AFM-SMFS data since the current algorithms cannot specify an unbinding force to the corresponding binding site due to the lack of networking functionality to model the relationship between the unbinding forces. 

"Because AFM-based measurements deal with nanoscale details of samples and thus are error-prone due to human and experimental conditions, the integration with AI is to overcome these shortcomings.," said Bingqian Xu, professor of electrical and computer engineering in the UGA College of Engineering and co-P.I. "This is crucial, especially for DNA-ligand interactions, as these interactions occur at localized sites (specific DNA target sequences), the atomic scale details of binding sites and the nature, strength, and stability of the interaction between DNA and ligands determine the properties and functions of the formed complex. " 

Image: Wenxuan Zhong