I have some feature matrixes from a bulk RNA seq analysis pipeline, and I want to perform bulk deconvolution to get relative single cell proportions. Some of the samples are from WT mice, while others are from mices with an induced condition.
I have also acquired a single cell dataset for the tissue and cell types ( around 40) that I want to analyze.
I have tried to use Music2 to perform this deconvolution, but the only times it converges are when I specify two or three cell types to get the proportions. When I specify more than that, it usually does not converge or it runs for 20+ hours without returning a result.
My question is, is there any similar tool to Music2 that does not require a single cell reference dataset for the disease induced samples? I can only find ones such as CIBERSORT that only work when both scRNA datasets are available.