From vegconomist.com
Researchers at the University of Leeds have built an artificial intelligence system that screens tens of millions of plant proteins to identify which ones can function as emulsifiers, narrowing the field to nearly 800 candidates. The tool is intended to replace years of manual laboratory testing with a computational shortcut for finding plant-based alternatives to dairy and synthetic emulsifiers used across food, cosmetics, and pharmaceutical products.
Emulsifiers keep oil and water blended in products ranging from mayonnaise and ice cream to lotions and medicinal creams. Many current options rely on animal-derived proteins such as casein and whey, or on synthetic compounds with a high carbon footprint.
A shortcut around trial and error
The research, published September 3 in Communications Chemistry, was led by postdoctoral researcher Dr Simha Sridharan and supervised by Professor Anwesha Sarkar, both based in the University of Leeds’ Sarkar Lab within the School of Food Science and Nutrition. The team worked with Dr Rik Sarkar, a machine learning specialist at the University of Edinburgh.
“There are millions of potential plant proteins, but testing them all to identify the right emulsifier is expensive and involves a time consuming trial and error approach,” said Sridharan. “Until now, there has been no reliable way to predict which plant proteins are likely to behave as emulsifiers like animal proteins.”
The team built a simulation model to study how proteins position themselves at the boundary between oil and water, then applied machine learning to identify the molecular segments, described as di-block structures, that determine this attachment behaviour.
“We were able to model this structure mathematically for plant proteins. Using machine learning based on features obtained from statistical physics simulations, we can predict which plant proteins are most likely to work best as natural emulsifiers,” said Rik Sarkar.
Pea and potato proteins confirmed in lab tests
Applying the model to tens of millions of candidate proteins narrowed the list to nearly 800 with emulsifying potential, many of which had not previously been studied for this application. The team then tested a selection of commercially available proteins against the model’s predictions, finding matches in proteins derived from peas and potatoes.
Anwesha Sarkar, who also co-directs the National Alternative Protein Innovation Centre (NAPIC) at Leeds, said the results point to a faster route for ingredient discovery. NAPIC has previously funded UK academic-industry collaborations in areas including fungal fermentation and school meal uptake studies, among other alternative protein projects.
The findings are aimed at food and cosmetics manufacturers seeking natural emulsifiers as consumer demand shifts away from animal-derived and synthetic options.
“This shows how AI could help researchers find promising new ingredients much faster than before,” said Sarkar.
No comments:
Post a Comment