Protein scientists and AI collide

Protein scientists and AI collide

A team of protein scientists from Rutgers University came up against a computer program.

Spoiler alert: the AI ​​won. But only by a hair.

Pairing humans with AI

The scientists decided they wanted to conduct an experiment that paired a human with a deep understanding of protein design and self-assembly with an artificially intelligent computer program with predictive capabilities. Topping the list of potential scientists was Vikas Nanda, a researcher at Rutgers’ Center for Advanced Biotechnology and Medicine (CABM).

The experiment aimed to determine whether the human or the AI ​​could do a better job of predicting which protein sequences would combine best.

The results were published in Natural chemistry.

Nanda, researchers at Argonne National Laboratory in Illinois, and various colleagues in the United States say the battle was “close but decisive”. The competition put Nanda and several colleagues against the AI ​​program, including one by a small margin.

Scientists are seeking more knowledge about protein self-assembly, believing that by understanding it better, they could design innovative new products for medical and industrial use. One of these products could be artificial human tissue for wounds while another could be a catalyst for new chemicals.

Nanda is a professor in the Department of Biochemistry and Molecular Biology at Rutgers Robert Wood Johnson Medical School.

“Despite our vast expertise, AI performed as well or better on multiple datasets, showing the enormous potential of machine learning to overcome human biases,” Nanda said.

Design and self-assembly of proteins

Proteins are made up of large numbers of amino acids linked end to end, and the chains fold together to form three-dimensional molecules with complex shapes. The shape of each protein and the amino acids it contains determine its behavior. Researchers like Nanda are involved in “protein design”, meaning they create sequences that produce new proteins. The team recently designed a synthetic protein capable of rapidly detecting VX, a dangerous nerve agent. This new development could have big implications for new biosensors and treatments.

Proteins self-assemble with other proteins to form important superstructures in biology. In some cases, proteins appear to follow a design, as in the case where they self-assemble into a protective outer envelope of a virus. Other times, they self-assemble when forming biological structures associated with certain diseases.

“Understanding protein self-assembly is fundamental to making progress in many fields, including medicine and industry,” Nanda said.

Nanda and five other colleagues were given a list of proteins and asked to predict which were likely to self-assemble. The predictions were then compared to those of the computer program.

Human experts used rules of thumb based on their observation of protein behavior in experiments, including patterns of electrical charges and degree of water aversion. They selected 11 proteins that they expected to self-assemble while the AI ​​selected nine proteins.

Their experiment showed that the humans made six correct predictions about the 11 proteins while the computer program picked nine.

The experiment also demonstrated that human experts “favored” certain amino acids over others, leading to incorrect choices. The AI ​​correctly picked out some proteins with qualities that didn’t make them obvious.

“We’re working to get a fundamental understanding of the chemical nature of the interactions that lead to self-assembly, so I was concerned that using these programs would miss important insights,” Nanda said. “But what I’m really starting to understand is that machine learning is just one tool among many, like any other.”

#Protein #scientists #collide

Leave a Comment

Your email address will not be published. Required fields are marked *