Alphabet (NASDAQ:GOOG) (NASDAQ:GOOGL) has expanded into healthcare on several fronts in recent years. There's now key progress in one of those areas: Google's DeepMind artificial intelligence (AI) software is being used to predict the structure of proteins. In this Motley Fool Live video recorded on Nov. 16, 2021, Fool contributors Keith Speights and Brian Orelli discuss whether or not DeepMind could even transform healthcare.

Keith Speights: Brian, let's totally switch gears here. We talked in the past about Google's DeepMind artificial intelligence software being used to predict the structure of proteins.

Now there's more information about the progress that's been made on this front. How is DeepMind being used, and is it an exaggeration to say that this has the potential to transform healthcare?

Brian Orelli: Yes. This new AI can predict protein interactions. Folding is essentially interactions. Proteins are a long chain of amino acids, and the backbone is for all the amino acids are the same, and those are the ones that interact with each other in the chain. But then they have different sides groups on them. Some are big, some are small, some are positive, some are negative, some like to interact with water so they are more likely to be on the outside of the protein because that's where the water is. If you look at two proteins, it's just the side groups of the proteins that are on the outside of the protein after it's folded that are interacting with each other.

Researchers started using the initial program. What they did was they took two proteins, that they wanted to see their interaction of, and they just told it that it was one protein. They took the sequence for the one protein, and then they put an artificial linker that they knew was flexible enough to just flop around and then it wouldn't create any structure to it, but it would allow the second half of the protein to fall back.

You could think protein and then linker, and then another protein and then these two proteins could interact with each other. They started doing that, and they could actually show that the one protein would then fold back on itself and interact with it. Then DeepMind said, "Let's just use that."

They developed this new AI that will look at the interaction of two different proteins. You don't have to do this artificially by creating this linker in one protein. I could see using this in a couple of different ways. You could use it to predict which proteins would interact with your protein of interest. That's more-basic science, although perhaps companies could use it to find a new target to use. You had a protein that you knew was involved in a disease, then you could go search out and you could use this protein to see what proteins are most likely to interact with it. Then you could use that new protein. You could then go target that new protein.

You could also in theory use it to figure out where the two proteins are interacting and then disrupt that interaction if that interaction is important for the disease progression or the cause of the disease.

Most drugs are small molecules, and they actually block the activity of an enzyme. This isn't really going to help with that development. I think the structure, the basic the first AI that allowed you to create the structure will help with those drugs. Some drugs help stabilize the protein.

Again, I think just knowing the general structure of the protein is where that would help with those kinds of drugs. There are some instances where you might want to block the interaction of two proteins. Knowing this would be helpful. Honestly, I think there's probably more helpful for basic science and I think it is for drug development. I think it could have some promise in drug development.

Speights: Brian, it sounds like you're saying not transform healthcare but an important advance.

Orelli: Yeah. I think it's an important advance in understanding biology, and so ultimately those things trickle down to drug development. But I see a limited usefulness in drug development at least right down.

This article represents the opinion of the writer, who may disagree with the “official” recommendation position of a Motley Fool premium advisory service. We’re motley! Questioning an investing thesis -- even one of our own -- helps us all think critically about investing and make decisions that help us become smarter, happier, and richer.