Error using stringtie merge

Question

[cwj6]

I'm trying to merge 6 gtf files I made with stringtie. I've written their file names in a .txt file in the proper format and named it mergelist.txt. The files I want to merge are in the directory I'm currently using. When I try to run:

stringtie --merge -p 4 -G /home/chris/Desktop/Mus_musculus_UCSC_mm10/Mus_musculus/UCSC/mm10/Annotation/Genes/genes.gtf -o stringtie_merged.gtf mergelist.txt

I get the error:

Error: cannot find file 'IP1t_stringtie.gtf' and mergelist.txt does not look like GFF!

The IP1t gtf is the first file in the list of 6 gtfs. All appear to be formatted validly if I inspect them, and they're straight out of stringtie itself. I'm sure I'm making a simple mistake, but I can't figure out what it is and googling these errors hasn't turned up anything. Thanks in advance for any help.

[GenoMax]

Can you try the following? Some programs are sensitive to the order of options given.

stringtie --merge -p 4 -o stringtie_merged.gtf -G /home/chris/Desktop/Mus_musculus_UCSC_mm10/Mus_musculus/UCSC/mm10/Annotation/Genes/genes.gtf  mergelist.txt

GTF files listed one per line in mergelist.txt, correct?

[cwj6]

Thanks for the reply. I got the same error after flipping around the -g and -o commands, unfortunately. And yeah, the gtf files are all on separate lines with no other formatting.

[GenoMax]

Can you try adding full path to the GTF file names in your mergelist.txt?

[cwj6]

Yup, I get the same error but with the full path instead of just the file name. Maybe I'll just start from the beginning and make all new gtf files.

Answer 1

[anngrenda]

Hello.

I am currently facing the exact same error. Any luck with your mergelist.txt yet? What was the problem after all?

[cwj6]

I never actually found a solution to this (or related) problems even after re-doing the analysis from the beginning. What I ended up doing was just bypassing the merge step and going straight to stringtie -eB just using the basic gene.gtf as the -G input, not a merged .gtf file. So for example:

stringtie -e -B -p 4 -G /home/chris/Desktop/tacr3_rnaseq/Genes/genes.gtf -o ballgown/i1/i1_stringtieout.gtf i1samout.bam

After analyzing with ballgown I got roughly similar outputs as I did with tophat/bowtie/cuffdiff, so I'm assuming that'll work OK.

Answer 2

[megan.chan]

Also having the same issue, I was able to merge successfully earlier this week with a different set of GTF files, but they were generated in the exact same way. This leads me to believe the problem is actually in the GTF files rather than mergelist.txt. Did you figure out why?

[megan.chan]

Solved my issue. I was using the wrong version of the reference file (-G).