Isoforms and downstream analysis

[aloliveira]

Hello,

I am very new in this transcriptome field and I have one question concerning the downstream analysis. I already assembled all my data, and now I'm preparing the data to the downstream analysis. I want to perform phylogenomics analysis, phylogenetic analysis with certain families of genes, comparative transcriptome analysis, and also functionally characterize all my transcriptomes. So, my initial question is:

For phylogenomics, phylogenetics, functional annotation is needed to pick just one isoform per component (gene)? I think to many isoforms could inflate the results (and I don't even know if the isoforms are real, or derived from sequencing errors, even that I performed a very aggressive filtering step). So, would be a good idea make a transcript quantification with RSEM and pick only the most abundant transcript per component based on some statistics? For example, fpkm?

I think I need to decide this question at this point, because it is crucial for the next steps. Thanks in advance,

Best regards,
André

[mchizar]

ArrayStar software provides both "gene level" and "isoform level" analysis of RNA-Seq assemblies and generates RPKM values at both annotation levels. This software provides a nice UI and is designed for people with limited bioinformatics experience. THere are free demos, too.