DESeq2 on a Novel Organism

Question

[pthom010]

I have a series of reads from an organism that has not been sequenced or annotated. I am using DESeq2 (through R) to do a differential expression analysis on my genes and was wondering if it was feasible to annotate them using a tx2gene file with Gene IDs in the third column? Theoretically, I would then be able to use NCBI Gene IDs in functional annotation software.

[Asaf]

If it's not annotated how could you have a tx2gene file? Can you please elaborate?

[pthom010]

I'm honestly very confused. Someone told me I could just customize the R scripts that Trinity uses but that would require a significant amount of customizing all of the R scripts (filtering of row means by genotype, specific FDR and FCs). I think it might be more efficient to run the analysis in DESeq2 by myself. My organism is novel (no annotations), so I hypothesized either mapping back to something with high levels of synteny or using NCBI Blast IDs as gene names to do GO annotation. I've read several papers with my organism explaining multiple methods and I've looked at archives of multiple questions and also have looked through multiple manuals and tutorials.

[Asaf]

You're mixing unrelated things. You have transcripts, that's great, and you probably have a count table, even better, you can now run DESeq2, I'm not familiar with the trinity R script but it shouldn't be too hard to run DESeq2 the "normal" way. Then you'll have differentially expressed genes and now you will want to understand what these genes are doing so you'll want to run some annotation software and that's the real challenge you're facing. There are some tools out there but they are domain specific so you'll have to tell us what is your "novel" organism to get suggestions (or you might find them yourself).