Hi Tej,

First before mapping the RNAseq reads to eukaryotic organism, first map the RNA-seq reads onto prokaryotic organism database. If you know the name of the that prokaryotic organism just download the respective database containing fasta sequences and map the RNA-seq reads to prokaryotic organism. Then take only those reads which are unaligned to prokarytic organi9sm and map it to the eukaryotic organism database and do your down stream analysis. This method you can be sure that you eliminated all the transcripts belongs to prokaryotic organism. But make sure you included all the genes from the that symbiotic prokaryotic organism

It might make sense to do a de novo assembly - I recommend Trinity - and then blast all the resulting sequences against either Uniprot (blastx, slow) or close relatives of both your organisms (blastn CDS/CDNA, blastx protein level).

How do you plan to do quantification without genomic sequences ? I believe genome free methods like Kallisto and Salmon also require a very good annotation / transcript set.

Hi Thamizh and Colindaven,

Thank you for the reply. I can not use simply similarity index to differentiate between them. The reason for that is, There are shared metabolic pathways between these two organisms. So if the gene is horizontally transferred then it will be the wrong idea to remove that gene sequence. I am considering this possibility because the gene that I am currently looking at is very significantly differentially expressed in the tissue. and showing similarity with the bacterial genome. Hence can not be ignored by just looking at the similarity between genome sequence of closely related species.

I am looking for a database /software that can differentiate between transcript based on their features e.g Poly-A tail is only present in eukaryotes. This information I can not use as in RNAseq data poly-A tail cold be missing. I hope you understand my problem.

Amol