Chemists discover AI's potential to combat food rancidity by predicting effective antioxidant combinations, revolutionizing food preservation.
A team of chemists seeking ways to extend the life of food products and combat the dreaded rancidity has made a significant breakthrough.
They have published a groundbreaking study in the Journal of Agricultural and Food Chemistry, highlighting the remarkable potential of artificial intelligence (AI) in preserving the freshness of food products containing oils and fats.
Rancidity and Antioxidants
The pesky issue of food going rancid occurs when it is exposed to air for an extended period, a process known as oxidation.
This report shared by Phys.org tells us that deterioration is particularly common in lipids, which encompass fats and oils. Heat and UV light can accelerate this process, forming ketones, aldehydes, and fatty acids, giving rancid foods that unmistakable, unpleasant odor.
While oxidation threatens food quality and, in some cases, consumer health, nature and the food industry have a potent defense against it-antioxidants.
These come in various forms, including natural molecules like vitamin C and synthetic ones that shield your food from the harmful effects of oxidation.
However, the challenge lies in choosing the right combination of antioxidants and calculating their precise amounts. This process is far from straightforward; adding antioxidants haphazardly can lead to decreased protective effects or antagonism.
Finding the right combinations for specific types of food involves numerous experiments, specialized personnel, and increased production costs, making it an arduous task.
AI Steps In
This is where AI enters the scene. Like ChatGPT, AI systems can process vast datasets, identify patterns, and generate valuable insights. The research team set out to teach an AI tool how to discover novel antioxidant combinations that would effectively combat food rancidity.
They chose an AI model capable of working with textual representations, which describe the chemical structure of each antioxidant.
To train the AI, the scientists initially provided it with a database of approximately one million chemical reactions, teaching it essential chemistry concepts such as molecular interactions.
As the AI grasped general chemical patterns, it was further fine-tuned using a database of over 1,100 antioxidant mixtures described in the research literature.
The AI quickly became proficient at predicting the outcomes of combining two or three antioxidants, aligning with the literature's descriptions 90% of the time.
However, real-world lab experiments did not always correspond with the AI's predictions, underlining the complexity of transferring results from a computer to the lab.
Refining the AI Model
AI models are not static; they are dynamic learners. The research team continued to feed the model new data, gradually sharpening its predictive abilities.
Adding about 200 examples from their lab experiments enabled the AI to accurately predict the effects of antioxidant combinations with minimal deviation from the actual outcomes.
The implications of this AI-powered approach are immense. It can assist scientists in developing more effective ways to preserve food by determining the best antioxidant combinations for specific food products, much like having a brilliant assistant in the lab.
Stay posted here at Tech Times.
Related Article: Shocking Study Reveals 1 in 7 Adults, 1 in 8 Children Addicted to Ultra-Processed Foods
Tech Times
