‘Deep fake’ built with AI could detect water pollutants
Image credit: T. Chris Gamblin
Researchers at the University of Kansas plan to use machine learning to create 'deep-fake' membrane beta-barrel proteins designed to detect polluting metal ions in water.
The machine learning process allows websites to scan millions of images on the internet to create fresh 'deep-fakes' could also be leveraged to detect water pollution, scientists say.
The team at the University of Kansas is looking at using a similar machine-learning process to generate a type of protein structure known as beta barrels that could be used in sensors to detect metal pollutants.
"These beta barrels are super useful because they can bring things across membranes," said principal investigator Joanna Slusky. "Barrels make good enzymes – there are so many different things that barrels can do."
Slusky and her co-principal investigators, professors Rachel Kolodny and Margarita Osadchy of Haifa University in Israel (along with KU postdoctoral fellow Daniel Montezano), will develop a new machine-learning process that generates beta-barrels with scaffolds similar to those found in nature, but with different sequences.
In the past, altering the binding properties of barrels required arduous work, that had to be done completely by hand and resulted in minor variations of a limited number of scaffolds, or barrel structures.
Instead, the scientists have chosen to use machine-learning processes to generate large numbers of barrels.
"There's a website called 'This X Does Not Exist'," Slusky said. "If you go to that site, you see all these AI-generated things and people don't really exist. But a computer made an image, for instance, of a cat. But that's not really a cat – a computer took a bunch of pictures of cats and said, 'OK, we can just sort of generate as many cat pictures as you want now, because we figured out what is a cat'. We need to make something real so we see it more like generating a recipe.
"The question is, how to make computers generate a recipe for proteins."
In a similar way to how a computer can generate that picture of a cat, Slusky's team believes it can train an algorithm to generate sequences that correspond to those of proteins that can be used for biosensors that would be able to identify pollutants like lead in waterways.
"If we make them the right size, this molecule will be ideal to put some particular metal in, and you can have the right substituents so that it would bind that metal," Slusky said.
"Because it's in a membrane, it can give you some sort of conductance difference – there's a difference between when it's bound and when it's not bound. If you're able to do that, you could sense for different metals, and different concentrations of those metals. There are a lot of big steps we want to accomplish, but I'm hopeful and excited."
In July, the UK Environment Agency called for water companies’ executives to face prison time if they oversee serious and repeated pollution incidents. In its latest annual report, the agency identified 62 “serious pollution incidents” that occurred last year, up from 44 the year before, in what it described as the “worst we have seen for years”.
That same month, a University of Portsmouth study revealed that the UK’s wastewater infrastructure is increasingly vulnerable to “pollution events” due to climate change, a prediction that the storms that hit the UK during the month of August seemed to confirm.
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.