Intron finder for 3' UTR sequences (non-model organism)?

Question

[LAB_2019]

I am trying to find introns retained in 3' UTR sequences. I have FASTA files for both the entire transcriptome & just the 3' UTR sequences for each transcript. Is there any software that can map introns using either FASTA file (preferably the 3' UTR.fa file)?

The GFF file I have contains annotated introns, but I believe they are exclusively in the coding regions. I have RNA-seq data that could also be used to search for introns in the 3' UTR if that is an option as well.

Thanks!

[Kristoffer Vitting-Seerup]

Is the transcritome fasta not just the nucleotide file with the sequence corresponding to the GFF - meaning it will contain the same regions (incl introns)?

Answer 1

[Kristoffer Vitting-Seerup]

Finding 3'UTR intron retention might be problematic due to the biological control mechanism called Nonsens Mediated Decay (NMD). Does that exist in your organism of interest? If so the analysis could be very hard to do.

That said I think your best approach would be to try and do a guided transcript reconstruction from the RNAseq data. Firstly will use tools designed to reliably identify new transcripts and then afterwards you could look for introns in the 3'UTRs (you would have to predict ORFs with other tools though).

You can read more about transcript reconstruction (incl tool suggestions) here. Also you can use my R package IsoformSwitchAnalyzeR to predict ORFs as well as integrate analysis of coding potential from tools such as CPAT and CPC2.

Answer 2

[LAB_2019]

Thanks for the advice. My organism does have NMD, so I do expect that 3' UTRs with introns would be degraded, but I'm hoping to still detect some (NMD is not 100% efficient).

In reference to your first question. The GFF file seems to have only introns annotated that were in the ORF, as I could not find any that mapped to the 3' UTR.

I will look into these tools and try to reconstruct the transcriptome with my data.