From this link remove X and Y chromosome genes in RNA-seq data using DESeq2 pipeline , I have learned that depending on context, it is perfectly valid to remove X and Y chromosomal genes in RNA-seq data before doing differential expression analysis. However I only have access to the count matrix and not the bam files of the RNA seq data I am analysing. Is it advisable to do the X and Y chromosome gene removal at the level of the count matrix? I am talking about removing all rows that contain X and Y chromosomal genes from the .csv count matrix file read in to R before proceeding to create the DESeq2 object.

Thank you in advance for your kind response.

Let me clarify that it's statistically valid. I don't know if it makes sense for your biologically question to do this.

If you have reason to remove X or Y genes, it is fine to do this on the count matrix, either the CSV or in R, before running DESeq().

This is perfectly normal. If you want to 'optimize' your non coding RNA analysis for example, you can also remove your mRNA genes.