How to distinguish level of expression in pooled-like samples?

Question

[ahmad mousavi]

Hi friends,

I have 2 question !!, suppose we have a pooled sample from different species (but same tissue), I know its uncommon but suppose organ transplantation or section transplantation.

My first question is for example we extract RNA from hamster and mouse heart tissue and pooled them, how could we distinguish those genes exactly expressed on hamster heart or mouse heart. I mean, although we could align them with both ref. genome separately but there is high similarity between both genome/transcriptome.

If we just rely on hamster/mouse specific genes we missed huge number of common genes. If we align without considering similarity we doubled their expression because of similarity. Is it any solution for such problems?

I have another question, If I make a pooled sample which contain ~100 mouse heart RNA samples and 1 hamster RNA sample, is it possible to see just expression of 1% of hamster genes due to low sample.

I think this pooling will decrease amount of hamster genes expression not number of expressed genes, am I right?

Thanks.

Answer 1

[h.mon]

From your post, it is not clear if you are talking about xenograft or transplants - as following quote implies:

pooled sample from different species (but same tissue), I know its uncommon but suppose organ transplantation or section transplantation.

Or if you are talking about pooling different species after RNA extraction:

we extract RNA from hamster and mouse heart tissue and pooled them

Xenografts are not that uncommon, and there are a number of specialized RNAseq tools: Xenome (but I couldn't find a download site), Disambiguate and BBSplit. BBSplit, which I used, can split your RNAseq reads into reads with unambiguous mapping for each parental genome, and reads with ambiguous mapping - map equally well to both genomes. If your genomes are sufficiently different, most of your reads will be assigned unambiguously to one parental genome.

I have another question, If I make a pooled sample which contain ~100 mouse heart RNA samples and 1 hamster RNA sample, is it possible to see just expression of 1% of hamster genes due to low sample.

This seems to me you want to pool different samples after extracting the RNA, I don't see the point of doing this.

[ahmad mousavi]

Hi and thanks for explanation. My question in Xenografts organ if the host tissue is much bigger than guest tissue/cell what happen then? (suppose 95% to 5%) I want to know in theory we missed some guest's gene expression or it reduce level of expression of guest tissue/cell??