rna-seq pipeline failed 'NoneType' object has no attribute 'chrom'

[kspham]

/usr/local/lib/R/site-library/DEXSeq/python_scripts/dexseq_count.py:169: UserWarning: Read FCC1Y2LACXX:2:1206:1071:92885#TCCTGAGC_CTCTCCCT claims to have an aligned mate that could not be found.
warnings.warn( "Read "+ i + " claims to have an aligned mate that could not be found." )
/usr/local/lib/R/site-library/DEXSeq/python_scripts/dexseq_count.py:169: UserWarning: Read FCC22FFACXX:8:1111:1391:11958#TCCTGAGC_CTCCCTTT claims to have an aligned mate that could not be found.
warnings.warn( "Read "+ i + " claims to have an aligned mate that could not be found." )
/usr/local/lib/R/site-library/DEXSeq/python_scripts/dexseq_count.py:169: UserWarning: Read FCC1Y2LACXX:2:1304:9951:45629#TCCTGAGC_CTCTCTTT claims to have an aligned mate that could not be found.
warnings.warn( "Read "+ i + " claims to have an aligned mate that could not be found." )
/usr/local/lib/R/site-library/DEXSeq/python_scripts/dexseq_count.py:169: UserWarning: Read FCC1Y2LACXX:2:1316:7855:80230#TCCTGAGC_CTCTCCTT claims to have an aligned mate that could not be found.
warnings.warn( "Read "+ i + " claims to have an aligned mate that could not be found." )
/usr/local/lib/R/site-library/DEXSeq/python_scripts/dexseq_count.py:169: UserWarning: Read FCC1Y2LACXX:2:2102:5292:11364#TCCTGAGC_CTCTCCGT claims to have an aligned mate that could not be found.
warnings.warn( "Read "+ i + " claims to have an aligned mate that could not be found." )
4000000 reads processed.
Traceback (most recent call last):
File "/usr/local/lib/R/site-library/DEXSeq/python_scripts/dexseq_count.py", line 239, in
if current_chromosome != a.iv.chrom:
AttributeError: 'NoneType' object has no attribute 'chrom'
' returned non-zero exit status 1
[2014-12-08 00:42] 3600000 reads processed.
[2014-12-08 00:42] Count reads in /media/workercc/htseq-count/tx/tmpmJfBAm/C008_Patch_2013-03-06_U01.counts mapping to /usr/local/share/bcbio-nextgen/genomes/Hsapiens/GRCh37ERCC/rnaseq/ref-transcripts.gtf using[2014-12-08 00:42]
[2014-12-08 00:42] ========== _____ _ _ ____ _____ ______ _____
[2014-12-08 00:42] ===== / | | | | _ | __ | | /\ | __ \
[2014-12-08 00:42] ===== | ( | | | | |) | |) | | / \ | | | |
[2014-12-08 00:42] ==== ___ | | | | _ <| _ /| | / /\ \ | | | |
[2014-12-08 00:42] ==== __) | |**| | |) | | \ | |** / ____ | |**| |
[2014-12-08 00:42] ========== |/ _/|**/|| _// _/
[2014-12-08 00:42] v1.4.4
[2014-12-08 00:42]
[

[kspham]

seems that we have a bug in deseq_count.py ? The latest (-u development) version got this error when running with ERCC.
File "/usr/local/lib/R/site-library/DEXSeq/python_scripts/dexseq_count.py", line 239, in

if current_chromosome != a.iv.chrom:

Thanks!

[chapmanb]

Son -- sorry for the delay. This is in the DEXSeq code base, not bcbio, so not exactly sure what is going on. @roryk do you understand that code at all to know what might be up?

[roryk]

Hi Son,

Sorry for the delay as well, I've been underwater and this skipped past me. It looks like something is borked about the alignment files, it looks like some reads are named "+ i + ". If you look through the alignment files, do they look corrupted somehow?

[kspham]

Yes !! They have + i +
Something is wrong with the alignment step (or ERCC preparation) I'm
digging into this.

On Mon, Dec 15, 2014 at 8:49 AM, Rory Kirchner notifications@github.com
wrote:

Hi Son,

Sorry for the delay as well, I've been underwater and this skipped past
me. It looks like something is borked about the alignment files, it looks
like some reads are named "+ i + ". If you look through the alignment
files, do they look corrupted somehow?

—
Reply to this email directly or view it on GitHub
#686 (comment)
.

[kspham]

ok, @chapmanb @roryk : The message :
warnings.warn( "Read "+ i + " claims to have an aligned mate that could not be found." )
was just the copied of the code that produce the warning. So I think it's no tthe problem.
The problem is in line 241: if (current_chromsome != a.iv.chrom):
here somehow a.iv is None , and that causes the problem.
But I still dont' understand why you switched to deseq_count instead of htseqcount?
The bamfile looks fine to be.

[roryk]

Hi Son,

Thanks for digging into the issue. dexseq_count is doing the exon level counts, so it doesn't replace htseq-count. Do you think you could put up the alignment file that is giving the issue somewhere so I can see if I can reproduce the error and figure out what is going on?

[kspham]

Definitely, will send you by email. (now)

On Mon, Dec 15, 2014 at 10:45 PM, Rory Kirchner notifications@github.com
wrote:

Hi Son,

Thanks for digging into the issue. dexseq_count is doing the exon level
counts, so it doesn't replace htseq-count. Do you think you could put up
the alignment file that is giving the issue somewhere so I can see if I can
reproduce the error and figure out what is going on?

—
Reply to this email directly or view it on GitHub
#686 (comment)
.

[roryk]

If this is blocking you from doing work if you install DEXseq, this step will get skipped.

[roryk]

uhh uninstall I mean. You can also move the dexseq.gff3 file somewhere else which will also skip it.

[kspham]

I never see the bcbio produces any deseq results in the final directory so
far -- Is it a new feature (that you recently added) towards diferential
spicing detection?

On Mon, Dec 15, 2014 at 10:48 PM, Rory Kirchner notifications@github.com
wrote:

If this is blocking you from doing work if you install DEXseq, this step
will get skipped.

—
Reply to this email directly or view it on GitHub
#686 (comment)
.

[roryk]

Hi Son, ironically, it has been in there a while but wasn't enabled while we made sure it worked properly. :)

[kspham]

ok -- I sent you the bam file :) could you access it ?

On Mon, Dec 15, 2014 at 11:00 PM, Rory Kirchner notifications@github.com
wrote:

Hi Son, ironically, it has been in there a while but wasn't enabled while
we made sure it worked properly. :)

—
Reply to this email directly or view it on GitHub
#686 (comment)
.

[roryk]

Yes-- thank you very much, it is downloading now. I'll take a look at it tomorrow and hopefully we can figure out what is wrong. Thanks!

[roryk]

I tested this file and it seems to be working okay for me. Is this the file that is giving the error? I see in the logs there are other files that are being run at the same time.

[kspham]

Hi Rory:
I think the source of the problem is the file ref-transcripts.dexseq.gff .
I sent you the gtf file (generated by bcbio script) in the email.

-Son.

On Tue, Dec 16, 2014 at 10:52 PM, Rory Kirchner notifications@github.com
wrote:

I tested this file and it seems to be working okay for me. Is this the
file that is giving the error? I see in the logs there are other files that
are being run at the same time.

—
Reply to this email directly or view it on GitHub
#686 (comment)
.

[roryk]

Thanks for sending me the files and working through debugging. Hm, I can't reproduce it with the files that you sent, it seems to be working okay. Could you stick the whole log file up as a gist to look at?

[roryk]

Hi @kspham,

Closing this out since it is pretty old-- I think things would be working okay. Let us know if it isn't working right. We have fixed a bunch of issues with the transcriptome generation using some odd GTF files since this happened so we might have ended up solving the issue. Please reopen if it is not working and we'll fix it for you. Thanks!