Advances in early human embryonic development and organogenesis from spatiotemporal multi-omics

Deciphering early human embryonic development and organogenesis is constrained by limited access to in vivo specimens and the intrinsic complexity of spatiotemporal cell fate dynamics. Recent advances in spatiotemporal multi-omics have begun to overcome these barriers by enabling the reconstruction of in silico human embryos and developmental trajectories with spatial and temporal resolution. These approaches have revealed fundamental principles of lineage specification, cell migration, tissue morphogenesis, and axial patterning in early human development. In this Review, we summarize how single-cell and spatial multi-omics technologies, particularly spatial transcriptomics, have transformed the study of human early embryogenesis, organogenesis, and stem-cell-derived embryoid systems. We highlight analytical frameworks that integrate spatial context, temporal progression, and multimodal datasets, and discuss how biological discovery and algorithmic innovation have co-evolved. Finally, we outline how artificial intelligence, foundation models, and large-scale multimodal learning may eventually enable predictive and causal modeling of human development, while addressing current technical, interpretability, and ethical challenges.