Single-cell based prognosis modeling identifies new breast cancer survival subtypes by cell-cell interactions

Description

This is the github repository for the project Single-cell based prognosis modeling identifies new breast cancer survival subtypes by cell-cell interactions by Shashank Yadav, Shu Zhou, Bing He and Lana Garmire et al.. It contains code and data for generating Figure 1-5 in the paper.

Getting Started

Dependencies

• Linux Working Environment
• Python 3
• R
• Anaconda3
• Jupyter
• Python libraries:
  • Numpy
  • pandas
  • Scipy
  • Scikit-learn
  • coxnnet
  • Theano
  • tqdm
  • pickle
  • seaborn
  • holoviews
  • plotly
  • lifelines
  • UnionCom
• R libraries:
  • NMF
  • circlize
  • BBmisc

Installing

Installing the R kernel on the jupyter

install.packages('IRkernel')
IRkernel::installspec()  # to register the kernel in the current R installation

Use the Bioconductor to install R packages.

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install(*PackageName* = "*Version*")

Use the package manager pip to install python packages.

pip install Numpy

Repository directories & files

The directories are as follows:

• 1_HMResult contains the heatmaps in Figure.1 of the manuscript.
• 2_PrognosisPred contains the modeling of prognosis prediction using single-cell phenotype features, constructing cox-nnet models
• 5_SankeyResult contains SankeyPlots descrbing how the NMF-defined classes intersect with the clinicopathological classification
• 5_AtypicalMatching contains class label transfer and comparison with TCGA-BRCA and METABRIC datasets.
• Figures contains Figure 1-7 showing in the paper.
• fig4_voilin/phen27_1 contains sub figures of Fig4.
• Pydata contains python data used for 5_sankey.ipynb and 4_Violin.ipynb
• rds contains .rds and .rdata data used for 1_HMOverview.ipynb, 3_NMFPlot.ipynb and 3_4_HMCircos.ipynb

The other files are as follows.

• 1_HMOverview.ipynb contains the ploting process of the general data heatmaps.
• 2_ProgPlot.ipynb contains the values in combining different sets of information from CP, TMI, and TCI.
• 3_NMFPlot.ipynb contains the result of NMF clustering and the consensusmap.
• 3_NMFScore.ipynb contains the visualization procedure of the silhouette and cophenetic score
• 3_kaplanmeier_OS.ipynb contains the kaplanmeier plot for NMF clustering and Grade, ER, PR, HER2 types.
• 3_4_HMCircos.ipynb contains the Heatmaps of Grade, ER, PR, HER2 and cell-cell interaction features for the NMF clusters and Circos plots demonstrate the correlation between features associated with each subpopulation. The export dimensions were enlarged to make Figure 3e. 4s annotations were put later in Adobe Photoshop for better explanation.
• 4_Violin.ipynbcontains scoring and profiling for the seven clusters based on various Cell phenotypes and Cell-Cell interaction features.
• 5_sankey.ipynbcontains procedures of constructing sankey plots

Local execution

• for .ipynb files: Using jupyter lab to execute the codes
• for .py files:

python3 -m *filename*.py

Authors

Contributors names and contact info

• @Lana Garmire
• @Shashank Yadav
• @Bing He
• @Shu Zhou

Current Maintainer

• Shu Zhou - https://github.com/Sukumaru

License

This project is licensed under the GNU General Public License v3.0 License - see the LICENSE.md file for details