The required packages and versions can be found in requirements.txt.
These notebooks contain codes for intensity normalization, nuclear segmentation, and saving cell image crops based on centroids.
processImgs.ipynb
processImgs_new_imgStats.ipynb
processImgs_new_stitchAndSave.ipynb
train_cnnvae.ipynb
This notebook performs the clustering and cluster validation process.
jointClustering/clusterIterate.ipynb
Notebooks starting with "train_allmeta_clf" contain codes for cross validation, using cluster proportion, neighborhood counts, or neighborhood compositions.
plot_ig.ipynb - Plots the integrated gradients of the clinical metadata classifications.
plot_imgFeatures_thresh8.ipynb
plot_clusterLocation_stats.ipynb
plot_clusterLocation_thresh8.ipynb
