A client sent us a kefir for full analysis. We isolated the community, sequenced what was there, and ran it against pathogen panels and metabolite screens.
We found a bacterial strain on the kefir with antimicrobial activity that was not on the label. The label came from supplier paperwork. No one had looked at the actual product at strain resolution.
We linked the genome to the function, validated the activity on pathogen panels, and confirmed the metabolite responsible with targeted MS. The client now owns the strain and the data that supports the claim.
Most labs would not have caught it, because most labs do not read microbiology, cell biology, and metabolic pathways in the same room. We do.