Fig. 5From: scNAT: a deep learning method for integrating paired single-cell RNA and T cell receptor sequencing profilesLatent space and GLIPH2 comparison on the MS dataset. A–C ResPAN (RNA) UMAP (left) and scNAT UMAP (right) colored by batch (A), source (B), and cell type (C). D, E Pair-wise normalized Euclidean distance comparison of the GLIPH2 clusters between RNA and scNAT in blood (D) and CSF (E)Back to article page