Cuffmerge Errors: cannot execute gtf_to_sam and other errors

[mhatha]

I am trying to run Cufflinks/cuffmerge/cuffdiff. So far I have aligned my RNAseq reads with STAR using the appropriate options to create a bam file that can be used in cufflinks. I then ran cufflinks with a gff file that I had modified and my bam files, one at a time, like this:

cufflinks -p 8 -G '/home/Mus_musculus.GRCm38.85_Modified.gff3' 'home/STAR_Alignments/Sample1Aligned.sortedByCoord.out.bam'

I did this cufflinks command for each bam file in my sample group.

Then I created a text file with the path to each output file from cufflinks, the transcripts.gtf files. Then tried to run cuffmerge like this:

cuffmerge -p 8 -g '/home/Mus_musculus.GRCm38.85_Modified.gff3' '/home/Cufflinks_Output/6_months/assembly_GTF_list.txt'

And all I can get is this error:
[Sun Sep 11 20:02:16 2016] Beginning transcriptome assembly merge
-------------------------------------------

[Sun Sep 11 20:02:16 2016] Preparing output location ./merged_asm/
[Sun Sep 11 20:02:29 2016] Converting GTF files to SAM
[20:02:29] Loading reference annotation.
GFF Error: duplicate/invalid 'transcript' feature ID=transcript:ENSMUST00000045689
[FAILED]
Error: could not execute gtf_to_sam


Can someone please help. I am not computer suave enough to figure it out on my own.

Thank you in advance.

[mhatha]

cufflinks is dropping the gene_id information

So in response to my earlier post I discovered that the cuffmerge program is not working because the transcript.gtf files that were generated by cufflinks are missing gene_id values for over 126,000 transcipts.

So for example this is what I am seeing in the transcript.gtf files created by cufflinks:
1 Cufflinks transcript 4774436 4774516 1 - . gene_id ""; transcript_id "transcript:ENSMUST00000045689"; FPKM "0.0000000000"; frac "0.000000"; conf_lo "0.000000"; conf_hi "0.097122"; cov "0.000000";
1 Cufflinks exon 4774436 4774516 1 - . gene_id ""; transcript_id "transcript:ENSMUST00000045689"; exon_number "1"; FPKM "0.0000000000"; frac "0.000000"; conf_lo "0.000000"; conf_hi "0.097122"; cov "0.000000";
1 Cufflinks transcript 4777525 4777648 1 - . gene_id ""; transcript_id "transcript:ENSMUST00000045689"; FPKM "0.0000000000"; frac "0.000000"; conf_lo "0.000000"; conf_hi "9.583444"; cov "0.000000";

I checked the GFF3 input file from ensemble and the gene id information is correct in that file. I modified the GFF3 file because I needed to add some isoforms that weren't in it, but that didn't cause the problem.

I am using the quick install static version of cufflinks that was downloaded and then I extracted the programs to the bin directory on my computer. Is there a bug in the program leading to this error? I've seen other similar posts on forums. Does anyone know that if I compile the cufflinks program from source if that will work better? Could that be the issue?

Thanks for any help.