Skip to main content
Fig. 9 | Genome Biology

Fig. 9

From: Normalization and de-noising of single-cell Hi-C data with BandNorm and scVI-3D

Fig. 9

Detection of two novel neonatal neuronal sub-clusters after BandNorm normalization. a Transformed p values from the comparison of scGAD scores of marker genes across the two sub-clusters. “New clusters” and “Original clusters” depict the sub-cluster labels obtained in this paper and in Tan et al. 2021 [11], respectively. The marker genes are inferred by Tan et al. 2021 [11] from the neonatal single-cell transcriptomics data. P values are obtained by Wilcoxon rank-sum test. b Heatmaps of gene expression of the marker genes with differential scGAD scores between the newly identified neonatal neuron I and II sub-clusters among the PV/SST interneuron, hippocampal pyramidal, and cortical L2-5 pyramidal neonatal cells. Marker genes with significantly different scGAD scores among the two sub-clusters are partitioned into two groups. Top heatmap: genes with significantly higher scGAD scores in the newly identified neonatal neuron I sub-cluster compared to neonatal neuron II sub-cluster. Bottom heatmap: genes with significantly higher scGAD scores in the newly identified neonatal neuron II sub-cluster compared to neonatal neuron I sub-cluster. c Gene ontology analysis of marker genes with significantly different scGAD scores between Neonatal Neuron I and II. d scGAD scores of lists of cell-type-specific marker genes, detected by comparing scGAD scores across cell types among single cells (left panel) and averaged for each cell type (right panel). scGAD score values are standardized into z-scores

Back to article page