Fig. 2
From: Complete deconvolution of cellular mixtures based on linearity of transcriptional signatures
Gene collinearity of bulk HNSCC TCGA data reveals pure cell types consistent with scRNA-seq data. a Regression lines and coefficients before and after normalization. Regression line coefficient will change to unity in case of true signature genes. b In case of true signature genes normalized vectors will be equal. c Workflow for TCGA samples analysis and validation with the scRNA-seq data. d Left: TCGA collinearity network after filtration: seven clusters of highly collinear genes and dark-gray cluster that contains small 2–3 genes clusters. e Single-cell data for HNSCC reanalyzed (GSE103322). f Expression profiles of genes in seven clusters in the scRNAseq dataset, color represents averaged Z-score for genes in cluster from blue (min) to red (max)
