scGCN is a graph convolutional networks algorithm for knowledge transfer in single cell omics

Abstract Single-cell omics is the fastest-growing type of genomics data in literature and public repositories. Leveraging growing repository labeled datasets transferring labels from existing to newly generated will empower exploration single-cell data. However, current label transfer methods have limited performance, largely due intrinsic heterogeneity among cell populations extrinsic differences between datasets. Here, we present a robust graph artificial intelligence model, Graph Convolutional Network (scGCN), achieve effective knowledge across disparate Through benchmarking with other on total 30 single datasets, scGCN consistently demonstrates superior accuracy leveraging cells different tissues, platforms, species, as well profiled at molecular layers. implemented an integrated workflow python software, which available https://github.com/QSong-github/scGCN .