Error with correctGCBias in deepTools

Question

[clmin]

Hello,

I'm trying to correct the GC bias with deepTools. I successfully used computeGCBias

computeGCBias -b file.bam --effectiveGenomeSize 2150570000 -g mm9.2bit -l 200 --GCbiasFrequenciesFile freq.txt -v

But the next step, correctGCBias, return an error I don't understand.

correctGCBias -b file.bam --effectiveGenomeSize 2150570000 -g mm9.2bit --GCbiasFrequenciesFile freq.txt -o gc_file.bam -v

applying correction

genome partition size for multiprocessing: 56882182

using region None

({'chrY': 'chrY', 'chrX': 'chrX', 'chr13': 'chr13', 'chr12': 'chr12', 'chr11': 'chr11', 'chr10': 'chr10', 'chr17': 'chr17', 'chr16': 'chr16', 'chr15': 'chr15', 'chr14': 'chr14', 'chr19': 'chr19', 'chr18': 'chr18', 'chrM': 'chrM', 'chr7': 'chr7', 'chr6': 'chr6', 'chr5': 'chr5', 'chr4': 'chr4', 'chr3': 'chr3', 'chr2': 'chr2', 'chr1': 'chr1', 'chr9': 'chr9', 'chr8': 'chr8'}, {'chrY': 'chrY', 'chrX': 'chrX', 'chr13': 'chr13', 'chr12': 'chr12', 'chr11': 'chr11', 'chr10': 'chr10', 'chr17': 'chr17', 'chr16': 'chr16', 'chr15': 'chr15', 'chr14': 'chr14', 'chr19': 'chr19', 'chr18': 'chr18', 'chrM': 'chrM', 'chr7': 'chr7', 'chr6': 'chr6', 'chr5': 'chr5', 'chr4': 'chr4', 'chr3': 'chr3', 'chr2': 'chr2', 'chr1': 'chr1', 'chr9': 'chr9', 'chr8': 'chr8'})

using 4 processors for 59 number of tasks

Sam for chr10 0 56882182

Sam for chr11 56882182 113764364

Sam for chr12 113764364 121257530

Sam for chr14 0 56882182

Sam for chr15 56882182 103494974

Traceback (most recent call last):

File "/usr/local/bin/correctGCBias", line 11, in <module>

main(args)

File "/usr/local/lib/python2.7/dist-packages/deeptools/correctGCBias.py", line 628, in main

writeCorrectedSam_wrapper, mp_args).get(9999999)

File "/usr/lib/python2.7/multiprocessing/pool.py", line 567, in get

raise self._value

ValueError: invalid value type 'B' (<type 'str'="">)

Could anyone help me?

[clmin]

I think that's all newest version:

deeptools 2.4.2-4-999ec5d

pysam==0.10.0

[Devon Ryan]

OK, please run this with one thread then and post the resulting error message. To properly fix this I might need a copy of the relevant portion of the BAM file (at least the region with the first alignment that's causing this behavior).

[Devon Ryan]

My guess is that this is an issue with the version of pysam that you're using. Can you post the following:

1. The version of deepTools you have installed
2. The version of pysam you have installed (pip freeze | grep pysam may tell you this).

I know it's a pain, but it might also help to run this single threaded. The error messages are more informative then.

[clmin]

I'm not sure I understand, my first post was one thread and the resulting error message. For the relevant part of the bam, I don't know what that it, sorry.

[Devon Ryan]

You were using 4 threads, even though you didn't specify it (it wasn't my idea, but deepTools will use 1/2 of the cores on a server unless you tell it explicitly how many it should use). Just rerun it with -p 1.

[clmin]

It seems the same:

correctGCBias -b ctr1.bam --effectiveGenomeSize 2150570000 -g mm9.2bit --GCbiasFrequenciesFile ct1_j28.txt -o gc_ct1_j28.bam -p 1 -v

applying correction

genome partition size for multiprocessing: 23672257

using region None

({'chrY': 'chrY', 'chrX': 'chrX', 'chr13': 'chr13', 'chr12': 'chr12', 'chr11': 'chr11', 'chr10': 'chr10', 'chr17': 'chr17', 'chr16': 'chr16', 'chr15': 'chr15', 'chr14': 'chr14', 'chr19': 'chr19', 'chr18': 'chr18', 'chrM': 'chrM', 'chr7': 'chr7', 'chr6': 'chr6', 'chr5': 'chr5', 'chr4': 'chr4', 'chr3': 'chr3', 'chr2': 'chr2', 'chr1': 'chr1', 'chr9': 'chr9', 'chr8': 'chr8'}, {'chrY': 'chrY', 'chrX': 'chrX', 'chr13': 'chr13', 'chr12': 'chr12', 'chr11': 'chr11', 'chr10': 'chr10', 'chr17': 'chr17', 'chr16': 'chr16', 'chr15': 'chr15', 'chr14': 'chr14', 'chr19': 'chr19', 'chr18': 'chr18', 'chrM': 'chrM', 'chr7': 'chr7', 'chr6': 'chr6', 'chr5': 'chr5', 'chr4': 'chr4', 'chr3': 'chr3', 'chr2': 'chr2', 'chr1': 'chr1', 'chr9': 'chr9', 'chr8': 'chr8'})

Sam for chr10 0 23672257

Traceback (most recent call last):

File "/usr/local/bin/correctGCBias", line 11, in <module>

main(args)

File "/usr/local/lib/python2.7/dist-packages/deeptools/correctGCBias.py", line 630, in main

res = list(map(writeCorrectedSam_wrapper, mp_args))

File "/usr/local/lib/python2.7/dist-packages/deeptools/correctGCBias.py", line 286, in writeCorrectedSam_wrapper

return writeCorrectedSam_worker(*args)

File "/usr/local/lib/python2.7/dist-packages/deeptools/correctGCBias.py", line 413, in writeCorrectedSam_worker

read.set_tags(readTag)

File "pysam/libcalignedsegment.pyx", line 2113, in pysam.libcalignedsegment.AlignedSegment.set_tags >

(pysam/libcalignedsegment.c:23066)

File "pysam/libcalignedsegment.pyx", line 295, in pysam.libcalignedsegment.packTags >

(pysam/libcalignedsegment.c:6217)

ValueError: invalid value type 'B' (<type 'str'="">)

[Devon Ryan]

OK, it's in the set_tags() function, which is the part I needed to know. Can you extract chr10:1-23672257 into a BAM file and post that and your freq.txt file somewhere? Then I can figure out a patch.

[clmin]

BAM file: https://e-nautia.com/share/45076-x8zr2q6p

Freqence file: https://e-nautia.com/share/45077-yzmr244h

[Devon Ryan]

Thanks, this turns out to be a bug in pysam. I'm trying to find out now if an earlier version handled this correctly. Whatever aligner you're using producing auxiliary tags that pysam is having problems with.

Answer 1

[Devon Ryan]

I've posted a bug report to pysam. You can follow the progress of it here.