cuteSV
cuteSV copied to clipboard
Installation fails
Hi, I tried to install cuteSV via pip and via the git repository, but it failed both times with an error involving PyVCF (see code snippet below).
It seems to be the same issue as described here: jamescasbon/PyVCF#332. In short PyVCF requires setuptools version <58
Someone mentioned in another thread (KarchinLab/open-cravat#98), that it seems that PyVCF only parses VCF 4.0 while 4.2 is the current standard.
Would it be possible to adjust cuteSV's dependencies on another VCF parser (also in regard to potential other compatibilities)? Thanks in advance.
Best, Paul
Running PyVCF-0.6.8/setup.py -q bdist_egg --dist-dir /tmp/easy_install-wrk_q1g2/PyVCF-0.6.8/egg-dist-tmp-yffxbp60
error: Setup script exited with error in PyVCF setup command: use_2to3 is invalid.
Hello, @pkerbs
Thanks for your valuable advice! In fact this dependencies brought inconvenience to the installation. There is an easier way to solve it by downgrading setuptools by pip install setuptools==57.5.0
.
In addition, we'll modify the dependencies to another VCF parser. The update will be released in later days.
Best, Shuqi
Hi Shuqi, thank you for your response. I tried that already, but unfortunately I still get the same error message:
× Getting requirements to build wheel did not run successfully.
│ exit code: 1
╰─> [1 lines of output]
error in PyVCF setup command: use_2to3 is invalid.
[end of output]
I am using a python virtual environment created using python 3.9.5 Do you have another idea what I am doing wrong?
Best, Paul
Hello, @pkerbs
I have created a virtual python environment through conda
, and I ran the installation by the following command:
# create the environment
conda create -n cutesv_env python=3.9.5
conda activate cutesv_env
# downgrade setuptools
pip install setuptools==57.5.0
# install cuteSV
pip install cuteSV
Through these, cuteSV can be successfully installed, and the pip list are shown as below:
$ pip list
biopython 1.81
cigar 0.1.3
cuteSV 2.1.0
Cython 3.0.6
joblib 1.3.2
numpy 1.26.2
pip 23.3.1
pysam 0.22.0
PyVCF 0.6.8
scikit-learn 1.3.2
scipy 1.11.4
setuptools 57.5.0
threadpoolctl 3.2.0
wheel 0.41.2
Just an idea, maybe you can try installing PyVCF v0.6.8 firstly? In fact, this dependency bring much inconvenience, and we are now fixing it in the next version.
Best, Shuqi
Hello Shuqi, I tried your suggestion and indeed it worked by installing PyVCF first in a freshly created virtual environment. What I did (with Python 3.9.5 installed):
python -m venv cutesv_venv
source cutesv_venv/bin/activate
pip install PyVCF
pip install cutesv
Thank you for working on a fix for the next version.
Best, Paul
Hello Paul,
That's nice! I'm glad to hear the success. I'll come back to this issue when the next version is released.
Best, Shuqi