CollasLab / edd

Enriched Domain Detector for ChIP-seq data
https://pypi.python.org/pypi/edd
MIT License
16 stars 4 forks source link

Error in reading bam file #12

Open salmonm opened 6 years ago

salmonm commented 6 years ago

Hello I'm trying to run EDD for the first time and I got the following error which I assume is related to reading the bam file Traceback (most recent call last): File "/opt/miniconda2/bin/edd", line 133, in import eddlib.experiment File "/opt/miniconda2/lib/python2.7/site-packages/eddlib/experiment.py", line 17, in import read_bam File "libcfaidx.pxd", line 51, in init eddlib.read_bam (eddlib/read_bam.c:8007) AttributeError: 'module' object has no attribute 'PersistentFastqProxy'

Any idea what is wrong? Thanks Mali

FatihSarigol commented 3 years ago

What's wrong is that a package/program version dependency problem prevents pysam from working which is required by EDD to read bam files by wrapping another program htslib. Not sure which installation method and which versions were used for this post 3 years ago, but the most recent bioconda build 1.1.19 also give the similar error.

I tried to install the 1.1.18 version too but that build was incomplete, missing some package dependencies.

I checked out the versions of the programs and noticed that in the current EDD conda build pysam is version 0.10.0 but that version was designed to wrap htslib version 1.3.2 while in the EDD conda build the htslib version is 1.9 and this causes pysam to not work apparently. https://pysam.readthedocs.io/en/latest/release.html#release-0-10-0

I first updated both to the most recent versions but that did not work, likely because this time some other old version packages caused dependency issue with the upgraded two.

I tried downgrading htslib to 1.3.2 and that worked.

For anyone that wants to use EDD today, the easiest way to install it is then via conda, first:

conda create -n EDD -c bioconda edd

then after installation is completed, downgrade htslib:

conda install htslib=1.3.2

Info about how to install bioconda: https://bioconda.github.io/user/install.html