Error while running DEXSeq

[Mudita]

I am working with Illumina RNAseq data. My experiment contain 5-5 biological replicates of normal and treated. Initially I followed cufflink-->cuffdiff workflow (Trimmomatics-->Tophat-->Cufflinks-->Cuffmerge-->Cuffdiff), but could not get any significant genes/isoform.

Later I started working working with HTSeq-->EdgeR workflow (Trimmomatics-->Tophat-->HTSeq-->EdgeR) and could get around 1000 significant genes.

Now I am looking for isoform information and so I started working DEXSeq package.

I am getting following error while running dexseq_count.py:

Traceback (most recent call last):
File "/data/RNASEQdata/Programs/dexseq_count.py", line 70, in <module>
for f in HTSeq.GFF_Reader( gff_file ):
File "/home/gadgil/.local/lib/python2.7/site-packages/HTSeq/__init__.py", line 210, in __iter__
for line in FileOrSequence.__iter__( self ):
File "/home/gadgil/.local/lib/python2.7/site-packages/HTSeq/__init__.py", line 51, in __iter__
lines = open( self.fos )
IOError: [Errno 2] No such file or directory: '-'

These are the steps I have followed:

python dexseq_prepare_annotation.py Homo_sapiens.GRCh37.71.gtf DEXSeq.gtf

samtools view accepted_hits.bam | python /dexseq_count.py -p no -s no -a 10 - DEXSeq.gtf output/19_N.txt

Please help me with this error.

Thanks in advance.

Best Regards,
Mudita

[areyes]

Hi @Mudita,

Have a look at the pasilla vignette, it contains examples of how to use the python scripts of the DEXSeq package: first you would have to convert the bam file into sam, sort it by read name and then run the python scripts.

Alternatively, have a look at the section "9.3 From GRanges, BamListFiles and transcriptDb objects" will indicate you how to generate an ExonCountSet object within R directly from bam files.

Alejandro

[Mudita]

Thanks for your reply

However, my bam file is already sorted by read name. Same kind of input i have already given to htseq and it worked there. The only difference of input I could see from HTSeq to DEXSeq, is the preperation of annotations file.

I downloaded annotation gtf file from ensemble:
http://asia.ensembl.org/info/data/ftp/index.html

I did this:
python dexseq_prepare_annotation.py Homo_sapiens.GRCh37.71.gtf DEXSeq.gtf

[areyes]

Hi @Mudita,

Sorry for my previous post (that I already deleted), it was incorrect!

I think the problem is the order of the parameters try this:

samtools view accepted_hits.bam | python /dexseq_count.py -p no -s no -a 10 DEXSeq.gtf - output/19_N.txt

[Mudita]

Thank you very much

it is working now. But it was just the matter of hyphen (-) before output file. I would have never figure it out.
Thanks again.

I would like to ask one more question here.
DEXSeq would only find out differentially expressed exons. However, I am also interested in splice junctions, which tool you would recommend other than cuffdiff.

Mudita