Stanford-led team reports cross-species cell AI models

Stanford Medicine researchers reported two AI models for cell biology, including TranscriptFormer, trained on data from 112 million cells across 12 species. The team says the model can compare cell types across species, identify cell types in species absent from training data, and distinguish healthy from diseased cells. Research papers on the work were recently published in Nature and Science. The source says treatment-design uses remain prospective.