This repository contains (often incomplete or broken) ideas for single cell analysis.
The minimal description of the working environment is:
xargs -L100 conda create -y -c aksarkar -c bioconda -c pytorch -n singlecell <<EOF
bioconductor-limma
cudatoolkit=10.1.243
colorcet
ipython
jupyter_console
matplotlib
numba
numpy
pandas
pyplink
pytest
python
pytorch
r-ashr
r-base
r-fasttopics
r-flashr
r-horseshoe
r-mrash
r-susier
scikit-learn
scipy
scmodes
scqtl
sphinx
tensorboard
torchvision
EOF
source activate singlecell
xargs -L100 python -m pip install <<EOF
-e /project2/mstephens/aksarkar/projects/anmf/
-e /project2/mstephens/aksarkar/projects/gtex-pred/
-e /project2/mstephens/aksarkar/projects/mpebpm/
-e /project2/mstephens/aksarkar/projects/nbmix/
-e /project2/mstephens/aksarkar/projects/poisbeta/
dca
git+https://github.com/aksarkar/scaa.git#egg=scaa
git+https://github.com/aksarkar/wlra.git#egg=wlra
h5py==2.10.0
hyperopt
leidenalg
pyro-ppl
rpy2==3.2.4
scanpy
scvi
EOF
We need to mix pip and conda due to incompatibilities in the biopython builds of scvi/scanpy.