This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Chromosome bias on RNA-Seq differential gene expression analysis

Hello everyone!

I conducted a differential gene expression analysis on a RNA-seq dataset to evaluate the alterations coming from the knockout of only one protein coding gene (comparison = knockout versus wildtype). Of a total of 88 upregulated genes, 73 are located on the same chromosome. I have already looked at the positions of the upregulated genes inside the chromosome and they are spread through all the chromosome. Besides, I checked for their functions and they don't seem to be functionally related. One possible explanation could be an extra copy of the chromosome (as the organism presents aneuploidy), but this is not an usual chromosome known to deviate from the diploid pattern. I'm waiting for a wet lab experiment to check ploidy anyway.

But, I would like to know if it could it be an artifact? If so, how can I check it?

Thank you!

chromosome-bias rna-seq

you expect to have aneuploidy only in the knockout?

It could be possible to observe an extra copy only in the knockout, it's a (pathogenic) plastic microorganism. But it's unlikely to happen in this case because we also have the add-back of the gene, which was made upon the knockout, and we didn't observe enrichment for that chromosome when comparing add-back versus wildtype. Thus, the extra copy would be gained in the knockout and lost in the add-back. I'll have an answer for this question in a while as we will run an experiment to check the number of copies of the chromosome in the knockout.

1 answer

If you look at all genes by chromosome, not just those that are nominally significant for differential expression, do you see an overall trend of increased expression across this chromosome? (You could probably do a permutation test for chromosome-level significance.) if so, this might indicate increased ploidy of that chromosome.

I highlighted all the genes from that chromosome in the volcano plot and, as you can see below, many genes are below threshold of p-value and/or fold-change. volcano_plot

Could you give more information on how to do a permutation test, please? Maybe a link...

Thank you!

Log in to answer this question.