nlesniak/deeplearning

Overview

project
|- README         		# the top level description of content (this doc)
|- CONTRIBUTING    		# instructions for how to contribute to your project
|- LICENSE         		# the license for this project
|
|- data/           		# raw and primary data, are not changed once created
| |- process/     		# .tsv and .csv files generated with main.py that runs the models
| |- baxter.0.03.subsample.shared      	# subsampled mothur generated file with OTUs from Marc Sze's analysis
| |- metadata.tsv     		        # metadata with clinical information from Marc Sze's analysis 		
|- code/          			# any programmatic code
| |- learning/    			# generalization performance of model
| |- testing/     			# building final model
|
|- results/        			# all output from workflows and analyses
| |- tables/      			# tables and .Rmd code of the tables to be rendered with kable in R
| |- figures/     			# graphs, likely designated for manuscript figures
|
|- submission/
| |- manuscript.Rmd 			# executable Rmarkdown for this study, if applicable
| |- manuscript.md 			# Markdown (GitHub) version of the *.Rmd file 
| |- manuscript.tex 			# TeX version of *.Rmd file 
| |- manuscript.pdf 			# PDF version of *.Rmd file 
| |- header.tex 			# LaTeX header file to format pdf version of manuscript 
| |- references.bib 			# BibTeX formatted references 
|
|- Makefile	 # Reproduce the manuscript, figures and tables

How to regenerate this repository

Dependencies and locations for Python code

• Python 3.6.5 or Python 2.7, Matplotlib, Numpy, Scipy, Sympy, Pandas, Sklearn and XGBoost to run Shallow Learning code.
• If running Deep Learning code you need to have Python 3 and Latest PyTorch and Latest Keras with Theano backend.
• Run everything from project directory.
• The files mentioned above at process/ from Marc Sze's analysis: https://github.com/SchlossLab/Sze_CRCMetaAnalysis_mBio_2018

Dependencies and locations for R code

• R version 3.5.0
• Caret package is used as the wrapper for all models
• Run from project directory.
• The files mentioned above at process/ from Marc Sze's analysis: https://github.com/SchlossLab/Sze_CRCMetaAnalysis_mBio_2018

Run the following code

git clone https://github.com/BTopcuoglu/DeepLearning

To run L2 Logistic Regression, L1 and L2 Linear SVM, RBF SVM, Decision Tree, Random Forest and XGBoost in Python

1. Generate tab-delimited files: Cross-validation and testing AUC scores of each model.
2. Generate tab-delimited files: The AUC scores of each hyper-parameter tested for each model.
3. Generate a comma-seperated file: The hyper-parameters tuned for each model in one file.
4. Generate ROC curve figures: The cross-validation and testing ROC curves for each model.

python code/learning/main.py

To run L2 Logistic Regression, L1 and L2 Linear SVM, RBF SVM, Decision Tree, Random Forest and XGBoost in R

1. Generate tab-delimited files: Cross-validation and testing AUC scores for each data-split and hyper-parameter AUC scores of each model.
2. Generate ROC curve figures: The testing ROC curves for each model.

R CMD BATCH code/learning/main.R

The Makefile will reproduce all the other figures and tables used in the manuscript.

make submission/manuscript.pdf