[BioC] Where is DEXSeq bioconductor package?

[Fabrice Tourre]

Simon,

Thanks.

I am curios what is the potential problem we will have if the -s
parameters set wrong? If there is a way to know the -s parameters from
the data?

When I run dexseq_count.py on my sam file from bwa, it always has this warning:

~/python/lib/python2.7/site-packages/HTSeq-0.5.3p1-py2.7-linux-x86_64.egg/HTSeq/__init__.py:543:
UserWarning: Malformed SAM line: RNAME != '*' although flag bit
&0x0004 set
  algnt = SAM_Alignment.from_SAM_line( line )
~/python/lib/python2.7/site-packages/HTSeq-0.5.3p1-py2.7-linux-x86_64.egg/HTSeq/__init__.py:592:
UserWarning: Read GA2-EAS337_0017_FC:1:8:11202:16260#0 claims to have
an aligned mate which could not be found. (Is the SAM file properly
sorted?)
  "which could not be found. (Is the SAM file properly sorted?)" )

but I can found GA2-EAS337_0017_FC:1:8:11202:16260#0 mate in the data,
as show in the attachments a.sam. And the a.sam is sorted by the name
(default by bwa).

In htseq-count, it also have this problem.

htseq-count a.sam Homo_sapiens.GRCh37.63/Homo_sapiens.GRCh37.63.gtf
100000 GFF lines processed.
200000 GFF lines processed.
...
2000000 GFF lines processed.
2064769 GFF lines processed.
Warning: Malformed SAM line: RNAME != '*' although flag bit &0x0004 set
Warning: Read GA2-EAS337_0017_FC:1:8:11202:16260#0 claims to have an
aligned mate which could not be found. (Is the SAM file properly
sorted?)
4 sam line pairs processed.

Why here report 4 line pairs? In fact, there are 3 sam line pairs in a.sam.

BTW: in your opinions, to do this kind of count using BWA or tophat
mapping, which one is better?  (Illumina pair-end human RNA-seq data)

Thank you.