Merge pull request #5 from kathyxchen/jupyter-notebook-ex-1

Notebook with PathCORE analysis workflow example

jupyter-notebooks/README.md

@@ -1,8 +1,24 @@
 # Description
 
-The two [Jupyter](http://jupyter.org/) notebooks in this directory
+Two [Jupyter](http://jupyter.org/) notebooks in this directory
 demonstrate the analyses that support Figure 3 and Supplementary Figure 2
-in the PathCORE paper.
+in the PathCORE paper:
+- [Figure3_overlap_correction.ipynb](Figure3_overlap_correction.ipynb)
+- [Supplemental2_TCGA_NMF_features.ipynb](Supplemental2_TCGA_NMF_features.ipynb)
 
 They must be run after `../ANALYSIS.sh`. Additional Python dependencies
-for the notebooks are specified in the `requirements.txt` file.  
+for the notebooks are specified in the `./requirements.txt` file for this
+directory.
+
+An additional notebook is included as a supplemental example of the PathCORE
+analysis workflow:
+- [Supplemental_PAO1_FastICA_example.ipynb](Supplemental_PAO1_FastICA_example.ipynb)
+
+In this notebook, we apply the PathCORE software to a FastICA model
+of the normalized _P. aeruginosa_ gene compendium. We apply
+[scikit-learn's FastICA implementation](http://scikit-learn.org/stable/modules/generated/sklearn.decomposition.fastica.html). 
+
+Similar to the case studies we describe in our manuscript, we set the number
+of features (ICA components) we construct to 300. A user interested in
+changing this parameter and examining the differences in the resulting
+network can do so in cell [4] of the notebook and re-run the analysis.