GATK conda install issues
Hi,
I'm having issues with installing gatk through conda on a HPC system. I originally installed gatk4 through https://anaconda.org/bioconda/gatk4 which worked fine until I got to the AnalyzeCovariates step of the Data pre-processing for variant discovery workflow where it errored because ggplot2 and gplots and potentually other packages were not installed. I installed ggplot2 through conda but gplots wouldn't install due to dependency issues.
So, I went back to install gatk using the approach shown in https://gatk.broadinstitute.org/hc/en-us/articles/360035889851--How-to-Install-and-use-Conda-for-GATK4#article-comments using gatk-4.5.0.0 but it wont create the envioment using conda env create -n gatk -f gatkcondaenv.yml. The command runs but ends with no environment made and this message:
Package libedit conflicts for:
r-base=3.6.2 -> krb5[version='>=1.16.4,<1.17.0a0'] -> libedit[version='>=3.1.20170329,<4.0a0|>=3.1.20181209,<3.2.0a0|>=3.1.20181209,<4.0a0']
conda-forge::python=3.6.10 -> sqlite[version='>=3.30.1,<4.0a0'] -> libedit[version='>=3.1.20181209,<3.2.0a0|>=3.1.20191231,<3.2.0a0']The following specifications were found to be incompatible with your system:
- feature:/linux-64::__glibc==2.34=0
- feature:|@/linux-64::__glibc==2.34=0
- bioconda::pysam=0.15.3 -> libgcc-ng[version='>=7.3.0'] -> __glibc[version='>=2.17']
- biopython=1.76 -> libgcc-ng[version='>=7.3.0'] -> __glibc[version='>=2.17']
- conda-forge::h5py=2.10.0 -> libgcc-ng[version='>=9.3.0'] -> __glibc[version='>=2.17']
- conda-forge::keras=2.2.4 -> tensorflow -> __cuda
- conda-forge::keras=2.2.4 -> tensorflow -> __glibc[version='>=2.17']
- conda-forge::mkl-service=2.3.0 -> libgcc-ng[version='>=7.5.0'] -> __glibc[version='>=2.17']
- conda-forge::numpy=1.17.5 -> libgcc-ng[version='>=9.3.0'] -> __glibc[version='>=2.17']
- conda-forge::pandas=1.0.3 -> libgcc-ng[version='>=7.3.0'] -> __glibc[version='>=2.17']
- conda-forge::python=3.6.10 -> libgcc-ng[version='>=7.3.0'] -> __glibc[version='>=2.17']
- conda-forge::scikit-learn=0.23.1 -> libgcc-ng[version='>=7.5.0'] -> __glibc[version='>=2.17']
- conda-forge::theano=1.0.4 -> libgcc-ng[version='>=7.3.0'] -> __glibc[version='>=2.17']
- r-backports=1.1.10 -> libgcc-ng[version='>=7.5.0'] -> __glibc[version='>=2.17']
- r-base=3.6.2 -> libgcc-ng[version='>=7.3.0'] -> __glibc[version='>=2.17']
- r-data.table=1.12.8 -> libgcc-ng[version='>=7.3.0'] -> __glibc[version='>=2.17']
- r-dplyr=0.8.5 -> libgcc-ng[version='>=7.3.0'] -> __glibc[version='>=2.17|>=2.17,<3.0.a0']
Your installed version is: 2.34
Note that strict channel priority may have removed packages required for satisfiability.
I read through similar posts about this but couldn't figure out if there was a way to fix this issue. So, does anyone have any recommendations?
-
It seems that this issue has just started to affect gatk conda environment installations especially a certain amount of time passes after a gatk stable release. I will check with our team to find a permanent solution to this issue. In the meantime you may want to use docker image for gatk which includes all the required dependencies in a single package.
I hope this solves your issue at least temporarily.
-
Thanks, I unfortunately can't currently use docker with the HPC system. But, I can try to find a solution to that if I can't fix the conda method.
-
Hi again.
Can you check the version of Conda you are using? Using very outdated versions may also cause this issue. Sometimes updating conda installation and rebuilding the environment from scratch can solve the issue. If not let us know.
Regards.
Please sign in to leave a comment.
3 comments