Tweet
Hi all,
I posted this in the RNAseq subforum but I realize that this forum might be more appropriate so I apologize in advance for the crosspost. I'm trying to map reads from my RNAseq experiment and I'm running into a strange problem. The command line I execute is:
tophat --bowtie1 -p 8 -N 1 -G genes.gff -o output genome t0.fastq
Tophat throws the following error and stops:
"Building Bowtie index from genes.fa
[FAILED]
Error: Couldn't build bowtie index with err = 1"
When I look at the tmp folder in the output directory, the genes.fa file is empty.
From the run log, the last thing tophat tried to execute was:
bowtie-build output/tmp/genes.fa output/tmp/genes
The previous command tophat tried to execute was gtf_to_fasta to make the genes.fa file from the provided gff file and no errors were reported in g2f.err despite genes.fa being empty.
If I don't provide a gff file, tophat runs just fine. I ran gffread -E on genes.gff and the gff file seems fine. I checked that the index names that bowtie-build generated from the genome file are the exact same as those in the gff file.
Strangely enough, I've been doing tophat analysis on other experiments that I've been running in the W303 strain of yeast. For this experiment I'm having trouble with, I'm using a different strain (SK1) and I've been using strain-appropriate genome and gff files (from the same source) and as far as I can tell, they are formatted correctly (both by visual inspection and by the above mentioned gffread). However, if I try to align these reads using the W303 genome and gff files I've been using for all my other analyses, everything works fine. Tophat only seems to have a problem with the SK1-specific files.
I'm very much puzzled at this point. Any help you guys could provide would be greatly appreciated! Thanks!
I posted this in the RNAseq subforum but I realize that this forum might be more appropriate so I apologize in advance for the crosspost. I'm trying to map reads from my RNAseq experiment and I'm running into a strange problem. The command line I execute is:
tophat --bowtie1 -p 8 -N 1 -G genes.gff -o output genome t0.fastq
Tophat throws the following error and stops:
"Building Bowtie index from genes.fa
[FAILED]
Error: Couldn't build bowtie index with err = 1"
When I look at the tmp folder in the output directory, the genes.fa file is empty.
From the run log, the last thing tophat tried to execute was:
bowtie-build output/tmp/genes.fa output/tmp/genes
The previous command tophat tried to execute was gtf_to_fasta to make the genes.fa file from the provided gff file and no errors were reported in g2f.err despite genes.fa being empty.
If I don't provide a gff file, tophat runs just fine. I ran gffread -E on genes.gff and the gff file seems fine. I checked that the index names that bowtie-build generated from the genome file are the exact same as those in the gff file.
Strangely enough, I've been doing tophat analysis on other experiments that I've been running in the W303 strain of yeast. For this experiment I'm having trouble with, I'm using a different strain (SK1) and I've been using strain-appropriate genome and gff files (from the same source) and as far as I can tell, they are formatted correctly (both by visual inspection and by the above mentioned gffread). However, if I try to align these reads using the W303 genome and gff files I've been using for all my other analyses, everything works fine. Tophat only seems to have a problem with the SK1-specific files.
I'm very much puzzled at this point. Any help you guys could provide would be greatly appreciated! Thanks!