multiBigwigSummary - Fatal Python error: deallocating None

180 views
Skip to first unread message

asa...@gmail.com

unread,
Jan 26, 2016, 9:37:19 AM1/26/16
to deepTools, andrew...@univ-amu.fr
Hi,
Using deeptools 2.0.1 (but same pb since 2.0.0), multiBigwigSummary fails for anything >4 bigwigs with the error:
Fatal Python error: deallocating None

if using multiple threads, this error is repeated and the script eventually fails. If using a single thread with -t 1, the error is thrown once followed by "Aborted"

Only happens with bigwigs and not bams.

reproduced with in-house bigwigs but also any encode bigwigs - e.g.:
ENCFF001LOU.bigWig
ENCFF001MVV.bigWig
ENCFF001MVW.bigWig
ENCFF001MVX.bigWig
ENCFF001NRI.bigWig
ENCFF001NYX.bigWig
ENCFF001NZP.bigWig
ENCFF001OBB.bigWig

command with 4 bigwigs which runs fine:
multiBigwigSummary bins -b BG/ENCFF001LOU.bigWig BG/ENCFF001MVW.bigWig BG/ENCFF001MVX.bigWig BG/ENCFF001NRI.bigWig -p 30 -o out.npz

gives: Number of bins found: 265500


add another bigwig and it fails:
multiBigwigSummary bins -b BG/ENCFF001LOU.bigWig BG/ENCFF001MVW.bigWig BG/ENCFF001MVX.bigWig BG/ENCFF001NRI.bigWig BG/ENCFF001NYX.bigWig -p 1 -o out.npz

gives:
Fatal Python error: deallocating None
Aborted

in case the added bigwig was the problem, included it in a group of 4:
multiBigwigSummary bins -b BG/ENCFF001MVW.bigWig BG/ENCFF001MVX.bigWig BG/ENCFF001NRI.bigWig BG/ENCFF001NYX.bigWig -p 30 -o out.npz

gives: Number of bins found: 265502


Any way to get around this? (or why it's failing?)

Thanks in advance

Andy

Devon Ryan

unread,
Jan 26, 2016, 9:52:24 AM1/26/16
to asa...@gmail.com, deepTools, andrew...@univ-amu.fr
I just tried 7 bigWig files and didn't run into any errors. Can you
make the bigWig files available to us? We can then debug things.

Devon
--
Devon Ryan, Ph.D.
Email: dpr...@dpryan.com
Data Manager/Bioinformatician
Max Planck Institute of Immunobiology and Epigenetics
Stübeweg 51
79108 Freiburg
Germany
> --
> You received this message because you are subscribed to the Google Groups "deepTools" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to deeptools+...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

Fidel Ramirez

unread,
Jan 26, 2016, 9:59:01 AM1/26/16
to Devon Ryan, asa...@gmail.com, deepTools, andrew...@univ-amu.fr
Also, check that you have installed the latest pyBigWig version (https://github.com/dpryan79/pyBigWig#installation).

asa...@gmail.com

unread,
Jan 26, 2016, 9:59:12 AM1/26/16
to deepTools, asa...@gmail.com, andrew...@univ-amu.fr, dpr...@dpryan.com

asa...@gmail.com

unread,
Jan 26, 2016, 10:11:54 AM1/26/16
to deepTools, dpr...@dpryan.com, asa...@gmail.com, andrew...@univ-amu.fr
Le mardi 26 janvier 2016 15:59:01 UTC+1, Fidel Ramirez a écrit :
> Also, check that you have installed the latest pyBigWig version (https://github.com/dpryan79/pyBigWig#installation).

Just upgraded from 0.2.1 to 0.2.3:

pip install git+git://github.com/dpryan79/pyBigWig.git --upgrade
Collecting git+git://github.com/dpryan79/pyBigWig.git
Cloning git://github.com/dpryan79/pyBigWig.git to /tmp/pip-VY7d2V-build
Installing collected packages: pyBigWig
Found existing installation: pyBigWig 0.2.3
Uninstalling pyBigWig-0.2.3:
Successfully uninstalled pyBigWig-0.2.3
Running setup.py install for pyBigWig
Successfully installed pyBigWig-0.2.3


But the error still throws...

Andy

Devon Ryan

unread,
Jan 26, 2016, 10:24:23 AM1/26/16
to asa...@gmail.com, deepTools
I can reproduce this, so I'll work on debugging the issue later tonight.

Devon
--
Devon Ryan, Ph.D.
Email: dpr...@dpryan.com
Data Manager/Bioinformatician
Max Planck Institute of Immunobiology and Epigenetics
Stübeweg 51
79108 Freiburg
Germany


asa...@gmail.com

unread,
Jan 26, 2016, 10:34:48 AM1/26/16
to deepTools, asa...@gmail.com, dpr...@dpryan.com
OK, I just answered that moving pyBigWig over to /usr/lib/python2.7/site-packages solved the problem, because it did.... once!

exact same command run again and error throws:

multiBigwigSummary bins -b BG/ENCFF001LOU.bigWig BG/ENCFF001MVW.bigWig BG/ENCFF001MVX.bigWig BG/ENCFF001NRI.bigWig BG/ENCFF001NYX.bigWig -p 30 -o out.npz

The following chromosome names did not match between the the bigwig files
chromosome length
chrM 16299
Number of bins found: 265500




multiBigwigSummary bins -b BG/ENCFF001LOU.bigWig BG/ENCFF001MVW.bigWig BG/ENCFF001MVX.bigWig BG/ENCFF001NRI.bigWig BG/ENCFF001NYX.bigWig -p 30 -o out.npz

The following chromosome names did not match between the the bigwig files
chromosome length
chrM 16299
Fatal Python error: deallocating None
<hangs until ctrl-c>

Devon Ryan

unread,
Jan 26, 2016, 3:54:53 PM1/26/16
to asa...@gmail.com, deepTools
Hi Andy,

Can you try the following and see if it fixes things:

pip install
git+https://github.com/dpryan79/pyBigWig.git@missing_pyincref --user
--upgrade

If that fixes things (it did for me) then I'll push out a new pyBigWig
release.

Devon
--
Devon Ryan, PhD
Bioinformatician / Data manager
Bioinformatics Core Facility
Max Planck Institute for Immunobiology and Epigenetics
Email: dpry...@gmail.com

asa...@gmail.com

unread,
Jan 27, 2016, 2:25:48 AM1/27/16
to deepTools, asa...@gmail.com
YES!
That fixed it - many thanks Devon for the speedy fix!
Andy


Le mardi 26 janvier 2016 21:54:53 UTC+1, Devon Ryan a écrit :
> Hi Andy,
>
> Can you try the following and see if it fixes things:
>
> pip install
> git+https://github.com/dpryan79/pyBigWig.git@missing_pyincref --user
> --upgrade
>
> If that fixes things (it did for me) then I'll push out a new pyBigWig
> release.
>
> Devon
>

Devon Ryan

unread,
Jan 27, 2016, 2:48:25 AM1/27/16
to A Saurin, deepTools
Great! I'll call that pyBigWig-0.2.4 then.

Devon
--
Devon Ryan, Ph.D.
Email: dpr...@dpryan.com
Data Manager/Bioinformatician
Max Planck Institute of Immunobiology and Epigenetics
Stübeweg 51
79108 Freiburg
Germany


Reply all
Reply to author
Forward
0 new messages