Inquiry About ReDeconv Workflow for Cell Deconvolution

[suhuanhou]

ReDeconv is an excellent tool, and I appreciate the developers for creating it. I would like to use this tool to deconvolve the cellular composition in bulk RNA-seq based on scRNA-seq data, but I haven't found any example tutorials.

My understanding of the steps is as follows:

1. Prepare the count matrix from scRNA-seq data.
2. Prepare the count matrix from bulk RNA-seq data (Can normalized matrices be used?).
3. Run the fourth function in ReDeconv_Normalization.py.
4. Execute all functions in ReDeconv_Percentage.py.

I am unsure whether this workflow is correct. I look forward to your reply.

[songjian2022]

Yes, the process is correct. The scRNA-seq data should be in raw counts, while the bulk RNA-seq data should be in TPM. If all cells in the scRNA-seq data originate from a single sequencing sample, the normalization step can be skipped.