Additional file 7 of Metabolic heterogeneity in clear cell renal cell carcinoma revealed by single-cell RNA sequencing and spatial transcriptomics

Additional file 7: Figure S7. Figures for scMet program evaluation. A Correlation between the number of training iterations for the Conditional Variational Auto-Encoder (CVAE) model and the corresponding training loss and validation loss. B Correlation between the number of cell-type specific markers used for deconvolution of RNA sequencing data and the accuracy of computational results (Left), and correlation between the number of cell-type specific markers used for deconvolution of RNA sequencing data and computational time (Right). C Bar plot representing the cell type proportions obtained after deconvolution of TCGA RNA-seq data. D Line graph illustrating the gene expression correlation between the best-fitted scRNA-seq data and the original RNA-seq data. E Scatter plot depicting the correlation between gene expression of TCGA-CJ-5684-01A-11R-1541-07 RNA-seq data and the best-fitted scRNA-seq data. F Workflow illustrating the utilization of small sample scRNA-seq data to convert eight TCGA RNA-seq datasets into scRNA-seq data using scMet. CVAE: Conditional Variational Auto-Encoder.