Micronoma, a San Diego-based biotech startup, was founded on the idea that signatures from microbes that colonize tumors could lead to a blood test to detect cancer. The hypothesis was based on a high-profile Nature paper, published in 2020, that suggested these microbes are consistently found in tumors, and that they create such specific signatures they could be used to identify cancer types with nearly 100% accuracy.
But late Monday night, a team of researchers published a manuscript on the preprint site bioRxiv.com showing what they believe to be two “fatal errors” in the Nature paper’s analyses, invalidating not only the paper’s findings but possibly the scientific foundation for Micronoma, which has raised an estimated $17.5 million in funding from investors including SymBiosis Capital Management and the Seerave Foundation.
One allegation is that the Nature paper’s authors didn’t adequately filter out human DNA from their database of tumor samples, leading them to incorrectly classify millions of human sequences as bacterial. The other is that a separate error led the team’s algorithm to find false associations between specific microbial signatures and cancer types.