This repository includes
- a minimal package for performing filtering
- all notebooks used to generate figures
See instructions below for installing and using the package.
pip install git+https://github.com/wanglab-broad/harmonics@main
This package can be used in Python as follows:
import scanpy as sc
import harmonics as hm
# Load data
adata = sc.read_h5ad('/path/to/your/data.h5ad')
# High-pass filter genes using a square-root (i.e. juxtacrine interaction) kernel
hm.sfilter(
adata,
'juxtacrine',
)
For a list of parameters, their meanings, and their types, run help(hm.sfilter)
.
Alternatively, one can run hm.sfilter
from the shell:
harmonics \
--read_path /path/to/your/data.h5ad \
--write_path /path/to/your/filtered_data.h5ad \
--filter_keys juxtacrine \
For a list of parameters in the shell, run harmonics --help
.