Is there a way to combine paired-end and single-read RNASeq for a genome annotation?

usegalaxy.eu support
1

Hi!

I’m using published data and have no way of generating new RNA-Seq data. There are two different sets of RNA-Seq data published for my non-model organism, one is paired-end reads and one is single-reads. I’m using HISAT2 to align and BRAKER3 to annotate. I’ve noticed that some annotations generated by the single library are missed from the paired library, and vice versa. Of course, most tools ask you to specify whether your data is single or paired, so these datasets aren’t compatible with each other.

Is there any way to combine these annotations in some way?

Thank you!

2

Hi @elpren

There is probably a way to do this once the data is has been reduced. You can also sometimes create different grouping levels where the sequencing method is factor but this depends on the tool.

What annotation files do you have right now and how do you want these combined?

3

Hi,

I have two GTF and GFF annotation files, one set from the paired library and one from the single library, both generated with BRAKER3. Ideally I’d like to combine the annotated features into one annotation file, but without duplicates that are annotated from both libraries. I’m not sure if this is possible at all, especially because lots of the attribute IDs are reused (they’re all just for example, g1.t1, g2.t1, g2.t2 etc.)

Currently I’m just viewing them as two different tracks with JBrowse, but I’d ideally like to do some DE analysis using the combined, more complete annotation.

Thanks for your help!

4

Hi @elpren

Thanks for explaining! I would suggest looking at the tool GffCompare.

You’ll be able to control how the merging is done. Maybe run a few times to decide on the best parameter choices for your goals?