Cafe: An integrated platform for exploring cell fate

Cell fate prediction from single-cell omics data is a central task for reconstructing developmental dynamics and inferring lineage relationships, but existing tools remain fragmented and often tailored to specific trajectory structures. Here we present Cafe (CelluAr Fate Explorer), an integrated platform for cell fate analysis that unifies trajectory-aware data storage, heterogeneous inference methods, benchmarking, visualization, and downstream interpretation. Cafe standardizes diverse inputs into a common milestone-network representation, supports both native and Dynverse-compatible methods, and evaluates performance across representative benchmark criteria. Applied to representative case studies across atlas-scale and lineage-resolved datasets, Cafe recovers coherent developmental trajectories and provides benchmark-guided method selection across data types and biological contexts. A browser-based interface built on cellxgene enables interactive trajectory inspection. Cafe also supports downstream analyses, including driver-gene prioritization and gene-regulatory-network inference. As an open, community-driven platform, Cafe is designed to evolve continuously with contributions from developers and users, expanding method coverage and functionality for cell fate prediction.