A study published on 20 October 2026 explores a possible route towards ingredients for foods with less sodium. The npj Science of Food paper reports an AI-guided search for fungal-derived peptides that produce salty responses or enhance saltiness in laboratory tests.

From computer screening to test solutions

Researchers screened candidates from 14 fungal proteomes, then synthesized six peptides. An electronic tongue identified leading candidates in water and a low-salt sodium chloride solution. Nature describes the publication as peer-reviewed accepted research, shared before its final edited version. This report draws on the accessible abstract.

A useful question for food development

The wider interest is practical: could an ingredient help developers reduce sodium while keeping a food appealing? WHO's sodium-reduction guidance, dated 11 May 2026, identifies reformulating products to contain less sodium as one public-health intervention. Product reformulation is the practical food-development question behind the laboratory work.

The next questions belong in real foods. Developers would want to understand how a candidate behaves in something such as soup, including its effects on bitterness, aftertaste and aroma. Cooking, storage and interactions with other ingredients would also matter. Evidence from finished foods and consumer testing would help answer those practical questions.

Keep the laboratory and the lunch separate

An electronic tongue measures responses through an instrument. It does not tell us, by itself, how people would describe or enjoy a meal. Computational screening can help prioritise candidates for investigation; performance in a finished recipe has its own demands.

For readers following food innovation, useful follow-up reporting would explain human taste testing, suitable food applications, safety assessment and feasible manufacturing. It would also show how much sodium a finished product actually contains. An appealing headline about flavour cannot answer those questions on its own.

For people interested in everyday meals, the lasting question is how a promising laboratory signal could become a useful food ingredient. The value of future findings will depend on evidence about the food people actually eat.

Original AI-assisted food-science reporting. Sources checked on 23 October 2026. Cover: AI-generated editorial illustration of ordinary home soup preparation, not the study's peptides, equipment or participants.

A good read is worth sharing.

Pass this story on to someone who would enjoy it.

Back to the journalPublished 23 October 2026 · Updated 24 October 2026

The conversation 0

A thought, a question, a little food for thought. Join in — no account needed.