Improving Gene Matching for CIBERSORT Deconvolution with Monaco Reference Matrix in Bulk RNA-Seq

Question

[chunglee13]

Dear all, I have a bulk RNA-seq dataset from a PBMC mixture that I want to deconvolute using the reference matrix from Monaco et al. However, I've noticed that only half of the genes in the Monaco reference matrix are present in my dataset. I used the AnnotationDbi package to compare Ensembl IDs between the two datasets.

1. How can I improve or fix the gene matching issue between my bulk RNA-seq data and the Monaco reference matrix?
2. How might these discrepancies affect the accuracy of the deconvolution using the CIBERSORT protocol?
3. Are there any recommended strategies or tools to handle missing genes or improve the overlap?

Any insights or suggestions would be greatly appreciated!