Notebooks for reproducing all figures and analysis of single-cell datasets for the "The Specious Art of Single-Cell Genomics" paper .
Where possible, analysis notebooks (.ipynb) are provided that can be run from Google Colab. Colab links are included at the top of the notebook. Just click on the symbol.
An intro to using Colab can be found here. Briefly, run each code cell by selecting the cell and executing Command/Ctrl+Enter. Code cells can be edited by simply clicking on the cell to start typing.
All saved/processed data used for analysis is streamed to the notebooks from CaltechData.
Github for Picasso Package: Package for running the Picasso algorithm and quantitative analysis on embeddings/matrices.
Github for MCML Package: Package for running the MCML algorithm and quantitative analysis on embeddings/matrices.
-
notebooks All analysis notebooks from which Figures were generated. Notebook sub-directories listed below for the corresponding Figures.
-
Figure 2:
jaccardDistort_Fig2a
,combinedFigurePlots_Figs2a_4a_7
,cellRankAnalysis_Figs2b_4b
,equidAnalysis_Fig2c
-
Figure 3:
uteroMixingAnalysis_Fig3a
,scanoCompMixingAnalysis_Fig3b
-
Figure 4:
celltypePred_Fig4a
,combinedFigurePlots_Figs2a_4a_7
,cellRankAnalysis_Figs2b_4b
-
Figure 5:
densityAnalysis_Fig5
-
Figure 6:
trajInference_Fig6
-
Figure 7:
picassoAnalysis_Fig7
,combinedFigurePlots_Figs2a_4a_7
-
Figure 8:
metricContrastAnalysis_Fig8
--- Supplementary Figures ---
-
Figure A,B:
jaccardDistort_Fig2a
,combinedFigurePlots_Figs2a_4a_7
-
Figure C-E,N-Q:
cellRankAnalysis_Figs2b_4b
-
Figure F-H:
equidAnalysis_Fig2c
-
Figure J,K:
uteroMixingAnalysis_Fig3a
,scanoCompMixingAnalysis_Fig3b
-
Figure L:
umapTransformAnalysis
-
Figure M:
celltypePred_Fig4a
,combinedFigurePlots_Figs2a_4a_7
-
Figure R,S:
densityAnalysis_Fig5
-
Figure T:
trajInference_Fig6
-
Figre U,V:
swissRoll
-
Figure W,X:
picassoAnalysis_Fig7
,combinedFigurePlots_Figs2a_4a_7
-
Figure Y,Z:
celltypePredMCML
,combinedFigurePlots_Figs2a_4a_7
-
Figure ZA :
bMCML
,combinedFigurePlots_Figs2a_4a_7
-
-
scripts
- Python scripts for Picasso and MCML algorithms and visualization functions.
- Python scripts for quantitative analysis of inter-/intra-distances and KNN metrics.
- R script for plots of PCA of equidistant points (Figure I in Supplement)
-
data
- Saved analyses outputs (csv).
-
env
- Conda environments (yml) for MACOS and Linux. Replicates Colab environment.