Entering edit mode
4.7 years ago
Dunois
★
2.9k
Let's say there are two species A and B. A and B are morphologically similar and are closely related (e.g., both belong to the same family).
Let's also say we have assembled a transcriptome from an individual that appears to fit the morphological descriptions of either of these species.
Is there any way one could confirm which of the two species this individual actually belongs to based on the transcriptome?
(Assume genome(s) and transcriptome(s) from prior experiments are available for the species A and B.)
That would depend on how complete the data it for the references and the quality of data you have at hand. If the species are very closely related then you may need to have complete sequence of 18S, 23S RNA to be certain. A prior paper for this type of work.
So if all data sets had the rRNA genes expressed, it would work? Thanks for the link, this will be useful.
If you are able to find full length sequence data that is of good quality I suppose so.
Hmm but there wouldn't be any rRNA if the RNA seq library was prepared by poly-A capture, correct? Except for residual "contamination" which need not necessarily be usable.
If the process worked perfectly then yes but you may still be able to find some.
Also, what if the transcriptome of the organism were to be mapped against the genomes of A and B? Could we look at the proportion of transcripts mapped successfully to each genome, and use that as a proxy for evolutionary distance? (I.e., more transcripts mapped to the genome of A than B, so the organism is more likely to belong to A.) I'm asking about this as an alternate solution given the event the rRNA genes are not available (for whatever reasons).
I am not an evolutionary biologist but that may be stretching things too far. Concept of a species becomes fluid especially with metagenomic samples (which is what I recall you have based on prior questions).
No, none of these samples are metagenomic. These are all marine eukaryotes that were sequenced from pooled individuals. And we're not even necessarily interested in assigning a species moniker to the unidentified organism. Just being able to tell if it's closer to A (or B) would be enough.
That is a much lower bar so may be achievable. You will need to see what your data looks like.