Hugging Face releases Carbon-A, a model it used to flag 566 million candidate genes
Hugging Face's biology team released Carbon-A and the Carbon Annotation Database, which it says contains 566 million new gene candidates across 22,617 species. Some candidates were validated in the lab with partners.
Hugging Face's HuggingFaceBio team announced Carbon-A on October 8, along with the Carbon Annotation Database built with it. The team says the model surfaced 566 million new gene candidates across 22,617 species, thousands of which had never been studied.
According to the post, a number of the new genes were validated in well-studied species such as cats, chicken and arabidopsis with partners at Active Site and UCSD, and more wet-lab results on previously unstudied genomes are due later.
The database, model, training data and technical report are published in a Hugging Face collection. The team frames the release as a way to keep biology research open rather than dependent on closed frontier models.