Hi folks,
I am working on comparative transcriptome of two related plant species, A and B. For each species, I have 3 control and 3 treated plants, and all samples contain reads between 15-25 millions. I have a genome and an annotation for species A, but not for species B. I finished Trinity assembly for species B, and I highly appreciate if you can throw light on my next moves, especially normalization and differential expression analysis. I am thinking of two approaches, but I am not sure which one is more appropriate.
1) I try using Transdecoder to predict a peptide for each transcript of species B, run Orthofinder (or similar) to find orthologs between species A and B, and use the information to create a master set of transcripts for the two species for abundance estimation/differential expression analysis. However, since Orthofinder is based on amino acid, I am not sure whether it will create a problem (in case of synonymous codons) when I try to map reads to the master transcripts for abundance estimation.
2) Skip the master transcript, I run DE analysis for each species individually and later use TransDecode/Orthofinder to find orthologs. This option sounds easier to me, but I don't know exactly whether it is possible to say 'an expression of gene XXX in species A is higher than in species B' since the normalization was done separately for each species.
Thank you very much in advance for all comments and suggestions.
What is the scientific question you are trying to answer? Comparing two species is akin to an apples to oranges comparison. You can do it but what scientific conclusion do you want to derive from it?
DE analysis is relative. You have a condition/set of samples (from the same organism) that is used as a control/baseline and you compare other set of treatments/time points against that condition calculating a relative change.
Hi genomax,
Thank you very much for your reply. :) I would like to explore how different the two plant species (in the same genus) respond to a pathogen since one species is much less susceptible to the pathogen than the other.
As you said that DE analysis is relative and for each species I am going to compare infected vs control plants anyways, would it make more sense (and easier) to follow option 2? May I ask if you have an alternative approach in your mind, please?
You did hold some critical information back in original post :-) So you have controls that you are going to be comparing against. Essentially these are two separate DE experiments in two species. You should be able to see if there are potential (qualitative) overlaps in the DE genes afterwards.
Start with number 2 and see how things go. I don't think you can quantitatively compare expression of any genes across species in this experiment. Perhaps qualitative differences will emerge and may explain resistance/susceptibility.
Oops, I am sorry. I should have mentioned that in the first post. :-( I will try with no.2 and hope things are going the right way. :)
Last question please, I am sorry for a basic question in advance. May I ask it is not possible to compare expression of any genes across species in this experiment, please? I thought I should have be able to compare (single copy) ortholog genes if I normalize the two species together.
Hi, did you compare your 2 species on a quantitative level in the end? I am learning about normalization which seems to be crucial in this case.