python -m pip install --index-url https://test.pypi.org/simple/ --no-deps scma==0.1.1

git clone https://github.com/tkisss/SCMA.git
cd SCMA
python setup.py install

scma.py -f INPUT_FILE -g group_name1 group_name2 group_name3... -o OUTPUT_PATH

INPUT_FILE: file of single-cell metabolomics, each two columns are a group which contains the ratio of nuclear to mass and its signal value. Cell names should be like groupname-xx, for example: A549-1,A549-2, gefitinib-6,gefitinib-8..., name before the first '-' will be considered as the group name, so the groups are A549 and gefitinib. Otherwise, -g should be provided.
OUTPUT_PATH: output path

• ppm_threshold_peak: peak error threshold for peak selection in the same cell. Default:10
• ppm_threshold_cell: peak error threshold when combining different cells. Default: 20
• decrease: peak selection. Default:True
• peak: whether the input file has selected peak. Default:False
• filter: peaks appearing in less than a% cells will be filtered. Default:0.5
• knn: use knn for missing value imputation. Default:True
• n_neighbors: knn algorithm parameter. Default:5
• method: dimensional reduction methods, including 'PLS','PCA','UMAP','tSNE'. Default:'PLS'
• p_value: differential analysis parameter. Default:0.05
• log2fold: differential analysis parameter. Default:0.5

Output will be saved in the output folder including:

• processed.csv: preprocessed data after peak selection and coordinate alignment
• filtered.csv: results after peak filtering, peaks appearing in less than a% cells will be filtered (parameter: --filter)
• knn.csv: results after knn imputation
• PLS.txt: dimensional reduction result
• de.csv: differential peaks among different groups (parameter: --p_value and --log2fold)
• violinplot.png: violin plot correspond to processed.csv, show how many cells each peak involved in
• embedding.pdf: dimensional reduction embedding
• heatmap.png: heatmap of differential peaks among different groups

Looking for more usages of scMA

scma.py --help