Extended Data Fig. 1: scVI’s latent space of Lawlor et al. data. | Nature Machine Intelligence

Extended Data Fig. 1: scVI’s latent space of Lawlor et al. data.

From: Iterative transfer learning with neural network for clustering and cell type classification in single-cell RNA-seq analysis

Extended Data Fig. 1

UMAP plot of scVI’s latent space when Baron human data were used as source data and Lawlor et al. data were used as target data. The plot indicates that scVI failed to remove batch effect between the source and target data, which led to low cell type annotation accuracy in the target data.

Back to article page