iss_patcher is a simple package for approximating features not experimentally captured in low-dimensional data based on related, high-dimensional data. The shared feature space between the two objects is identified, and log-normalised and z-scored on a per-object basis. The nearest neighbours of the low-dimensional observations in the high-dimensional space are identified, and the counts of the absent features are approximated as the mean of the high-dimensional neighbours.
While the function was initially written for processing ISS and GEX data, it can in principle be used for any sort of low-dimensional data featuring a subset of features from high-dimensional data.
show requirements
iss_patcher can run on a standard computer with enough RAM to hold the used datasets in memory.
OS requirements
The package has been tested on:
- macOS Monterey (12.6.7)
- Linux: Ubuntu 18.04.6 bionic
Python requirements
A python version >=3.7 and <3.12 is required for all dependencies to work.
Various python libraries are used, listed in pyproject.toml, including the python scientific stack with scipy>=1.6.0, annoy and scanpy.
iss_patcher and all dependencies can be installed via pip (see below).
Optional: create and activate a new conda environment (with python<3.12):
mamba create -n iss_patcher "python<3.12"
mamba activate iss_patcherfrom github
pip install git+https://github.com/Teichlab/iss_patcher.git(installation time: around 2 min)
Please refer to the demo notebook. Docstrings detailing the arguments of the various functions can be accessed at ReadTheDocs.
(demo running time: around 10 min)
iss_patcher is part of the forthcoming manuscript "A multiomic atlas of human early skeletal development" by To, Fei, Pett et al. Stay tuned for details!