DAta-driven Tools for Single-cell analysis using Image-Guided MAss spectrometry
This is the code repository containing the data-driven and machine learning based framework for image-guided single-cell MS data processing and interpretation, described in this paper: https://pubs.acs.org/doi/10.1021/acs.jproteome.2c00714.
- Numpy >=1.19.2
- Scipy >=1.6.2
- sklearn >=1.1.2
- PyTorch ==1.4.0
- Spike (FTMS processing): https://github.com/spike-project/spike
- First nevigate into the directory:
cd DATSIGMA - Create conda virtual env:
conda env create -f environment.yml - Activate virtual env:
conda activate datsigma - Install Jupyter Notebook:
conda install -c anaconda ipykernel - Add virtual env to kernel:
python -m ipykernel install --user --name=datsigma
The repository contains:
- Signal, image, and MS data preprocessing modules.
- Unsupervised analysis modules.
- Machine learning modules.
- processing 30,000 single cells raw high-resolution MS data
- exploratory analysis of neurons coupled with immunostaining
- machine learning classification and feature selection for neurons vs. astrocytes
- single vesicle classifications
- analysis of cells of developing brain
- 20,000 single aplysia neurons from six types of ganglia
Raw high-resolution FTMS data are available upon request due to large size. Processed data sets are available at Illinois Data Bank