scGPT: A Foundation Model for Single-Cell Multi-Omics

Researchers have developed scGPT, a generative pretrained transformer model trained on over 33 million single-cell sequences, to advance single-cell biology. By drawing parallels between language and cellular biology, scGPT effectively captures essential biological insights about genes and cells. The model demonstrates superior performance in various applications, including cell type annotation, multi-batch integration, multi-omic integration, perturbation response prediction, and gene network inference.