A platform for inference and analysis of microbial association networks. This platform provides tools optimized for noisy networks and includes all the functionality you need to start using Neo4j for your biological data. Native graph storage is the natural match for linked biological data!

Currently, the following features are available:

• Preprocessing of count files for network inference
• Clustering + splitting files by sample properties
• Batch network inference
• Network + BIOM file storage in a graph database
• Multi-network logic operations
• Taxonomy-dependent edge agglomeration More features are currently under construction.

Welcome to massoc! Contact the author at lisa.rottjers (at) kuleuven.be. Your feedback is much appreciated! This is version 0.3.0, and therefore still in early alpha. Encountering bugs is highly likely!

You can access massoc through a command line interface (CLI) and through a graphical interface (GUI). For tutorials with these two interfaces and a complete manual, check out the docs. The tutorials show you how to use the entire massoc pipeline, from BIOM file to network analysis!

To run massoc on Windows or Ubuntu, you only need to run the appropriate executable for your platform. These executables are stand-alone and do not need the other files in the repository. However, some of the features require you to have installed specific software, for example Neo4j Server. Read the sections below to figure out which dependencies you need to install and how to do so. Find the latest releases here.

If you do not want to use these executables or have macOS system, you can install massoc directly from the repository. This can result in some issues with dependencies.

pip install git+https://github.com/ramellose/massoc.git

There is no executable available for macOS systems, but you can set up your macOS system to run massoc as follows. First install macports, then run:

sudo port install python35
sudo port select --set python python35
sudo port install py35-pip
sudo port select --set pip pip35
pip install numpy==1.13.1 scipy h5py nose Cython psutil neo4j matplotlib networkx multiprocess sklearn scikit-bio biom-format --user
pip install -U wxPython --user

To reset to the default Python version, run:

sudo port select --set python python37
sudo port select --set pip pip37

To run network inference, you will need to ensure that your system can run your chosen network inference tools. If you want to run CoNet, SparCC or SPIEC-EASI from massoc, the instructions below will help you get started.

To run SPIEC-EASI from massoc, you will need the following:

• Rscript
• biomformat
• docopt
• SpiecEasi

The required libraries can be installed in your R environment with the script below.

source("https://bioconductor.org/biocLite.R")
biocLite("biomformat")
library(devtools)
install_github("zdk123/SpiecEasi")
install_packages("docopt")

Download the CoNet file from here. massoc will ask you for the location of the CoNet3 folder.

Download the SparCC file from here. massoc will ask you for the location of the SparCC folder. To run SparCC, you will need to have Python 2.6 or 2.7 installed.

On Windows, it will be necessary to set your PATH so both installations can be found. Find your Python 2 installation, rename the python.exe file to python2.exe and add it to your PATH.

The graph database utilities are built on Neo4j. Download the Neo4j Server community edition from here.

Extract the files from the distribution. On Windows, run bin\neo4j.bat console to initialize the server.

On Linux, first set your default Java version to 1.8.0:

sudo apt install openjdk-8-jk
sudo update-java-alternatives --set java-1.8.0-openjdk-amd64

For macOS, first install the appropriate Java version:

brew cask install java8

After downloading and extracting the Neo4j Server folder, initialize the server with bin\neo4j console.

For development purposes, the GUI can be executed from massocGUI.py.

massoc runs on Python 3.5. All its dependencies are listed in the requirements.txt file.

This software is still in early alpha. Any feedback or bug reports will be much appreciated!

• Lisa Röttjers - ramellose

See also the list of contributors who participated in this project.

This project is licensed under the Apache License - see the LICENSE.md file for details

massoc is largely built on and inspired by the BIOM format.