STAR genome indexing with multiple gff3 files

Question

[guylab]

Hello all.

I am trying to align some Pro-seq reads from mouse samples to the the mm10 genome using STAR. I'm mainly concerned with how these align to the mouse rDNA sequence, but thought it best to align to the rest of the genome as well, and then seperate out my region of interest from there.

To start with I was going to build a STAR index by supplying it with the separate .fasta and .gff3 files for each chromosome as well as the rDNA repeat sequence.

My problem is, although I can supply multiple .fasta files, I'm unsure of how to handle supplying the .gff3 files for each, as all the examples I have seen so far only supply one .gff file.

Has anyone else indexed mm10 in a similar fashion successfully?

Or alternatively, would it be better to append this rDNA sequence information to a comprehensive genome .fasta and .gff file and then index?

---

UPDATE: I decided to forgo STAR, and instead build a bowtie2 index using the chromosome's and rDNA .fa files to align to.

Answer 1

[Devon Ryan]

The answer to this is to either concatenate the gff files or just download one that has everything in it. It's very atypical to download files for only one chromosome, just avoid doing that in the future as they're largely useless.