polyactis / Accucopy

Accucopy is a computational method that infers Allele-Specific Copy Number alterations from low-coverage low-purity tumor sequencing data.
https://www.yfish.org/software/Accucopy
GNU General Public License v3.0
15 stars 4 forks source link

maestre: libbz2.so.1.0 not found #16

Closed menzel closed 2 years ago

menzel commented 2 years ago

maestre does not find the libbz2.so.1.0 and fails with the usual lib not found error message:

/opt/accucopy/maestre: error while loading shared libraries: libbz2.so.1.0: cannot open shared object file: No such file or directory

Even though I installed libbz2 for the current conda environment as well as on the system:

conda install -c conda-forge bzip2 yum install bzip2-libs

fanxinping commented 2 years ago

Hi, I think the reason is that you have installed bzip shared library in the conda env but it hasn't been added to the shared library search path. You can find where the libbz2.so.1.0 is and add it to the env variable LD_LIBRARY_PATH. For example: I find libbz2.so.1.0 in /y/home/fanxp/.local/anaconda3/envs/deep/lib, and I have two method to add it 1.LD_LIBRARY_PATH=/y/home/fanxp/.local/anaconda3/envs/deep/lib:$LD_LIBRARY_PATH Accucopy/main.py .......

  1. export LD_LIBRARY_PATH=/y/home/fanxp/.local/anaconda3/envs/deep/lib:$LD_LIBRARY_PATH
    Accucopy/main.py .......

    Method 1 is recommended and try mehtod 2 if method 1 is not work. Method 2 will pollute the shared library search path in the current user's shell, and some dynamically linked programs will failed if some shared library in the /y/home/fanxp/.local/anaconda3/envs/deep/lib conflict with those. If that occur, please logout and login to refresh LD_LIBRARY_PATH to default. Besides, using -Wl,-rpath flag when compile will store libbz2.so.1.0 searching path into the program and LD_LIBRARY_PATH is no longer needed if you familiar with compiler flag and Makefile.

Accucopy is hard to compile and using Docker image or Singulairty image is recommended.

menzel commented 2 years ago

Thanks for the fast replay and fixing my conda issues. I now have the next problem with GADA and libhdf5_cpp.so.100

I'll try the docker image later. Unfortunately, your domain yfish is marked as suspicious in our network and I cannot access it from work.