This is the repository for ADAGE (Analysis using Denoising Autoencoders for Gene Expression)
Python R Shell
Switch branches/tags
Nothing to show
Clone or download
tj8901nm Merge pull request #3 from greenelab/rrosh-patch-1
Update zero_one_normalization.py
Latest commit 2575a60 May 26, 2016
Permalink
Failed to load latest commit information.
Data_collection_processing Update zero_one_normalization.py May 25, 2016
Gene-gene_function first commit for adage Nov 3, 2015
Genome_organization first commit for adage Nov 3, 2015
Node_interpretation first commit for adage Nov 3, 2015
Train_test_DAs first commit for adage Nov 3, 2015
LICENSE first commit for adage Nov 3, 2015
README.md add info about published paper. May 17, 2016
pseudomonas_autoencoders.sh

README.md

adage

DOI

This is the repository for ADAGE (Analysis using Denoising Autoencoders for Gene Expression)

This repository provides the source code in support of the manuscript: ADAGE-Based Integration of Publicly Available Pseudomonas aeruginosa Gene Expression Data with Denoising Autoencoders Illuminates Microbe-Host Interactions. J Tan, JH Hammond, DA Hogan, CS Greene. mSystems, 00025-15.

############################################################

To set up ADAGE, first clone the repository. This is a short summary. Detailed instructions and steps to generate the model and reproduce analyses used in the manuscript are in pseudomonas_autoencoder.sh

Building an ADAGE model requires installing python packages Theano and Docopt Instructions for Theano: http://deeplearning.net/software/theano/install.html Instructions for docopt: https://pypi.python.org/pypi/docopt

We provide a gene expression compendium of Pseudomonas aeruginosa that contains datasets available before 02.22.2014. To get an up-to-date compendium, follow the instructions in Section One in pseudomonas_autoencoder.sh

Before training, first 0-1 normalize the compendium, run python Data_collection_processing/zero_one_normalization.py Data_collection_processing/Pa_compendium_02.22.2014.pcl Train_test_DAs/train_set_normalized.pcl None

To train a denoising autoencoders, run python Train_test_DAs/SdA_train.py Train_test_DAs/train_set_normalized.pcl --parameters

To test a dataset on an ADAGE model, run python Train_test_DAs/SdA_test.py Train_test_DAs/Genome-hybs_normalized.pcl --parameters

############################################################

Please email jie.tan.gr@dartmouth.edu if you have questions.