AI identifies 800 plant proteins to replace synthetic emulsifiers

AN AI model was able to identify nearly 800 plant proteins that could be used to replace synthetic emulsifiers in food and pharmaceuticals, according to a study from the University of Leeds.

Researchers in the School of Food Science and Nutrition were able to pin down a range of proteins that had not been considered previously in other research trials.

Emulsifiers are a core component of many everyday products, such as mayonnaise, make-up and ointments. They are needed to help mix ingredients that usually do not mix together, typically oil and water-based molecules.

Natural emulsifiers

In manufacturing, emulsifiers can either be natural, in the case of xanthan gum and lecithin, or synthetic, the most popular being DATEM.

There has been a growing interest in developing more natural emulsifiers with bioactive ingredients that may be more circular and biodegradable.

Lead researcher of the Leeds study Simha Sridharan and supervisor Anwesha Sarkar used AI and machine learning technologies to speed up the process of protein discovery. Using simulation modelling technology, the team learned how proteins interact with oil and water interfaces, and highlighted what characteristics of proteins influenced emulsions.

By collating this data, the researchers were able to predict what plant proteins were most likely to demonstrate emulsification properties when compared with animal-based proteins.

Protein power

To test the predictability of their experimental results, the researchers trialled several commercially available proteins, including pea and potato proteins.

The results were “promising”, with the proteins showing effective emulsification properties and supporting the predictions developed by the AI model.

Sridharan’s team say that the study shows that AI can be used as a reliable and supplementary tool for lab testing, providing a larger pool of potential protein sources to consider.