Skip to content

Latest commit

 

History

History
121 lines (89 loc) · 8.89 KB

README.md

File metadata and controls

121 lines (89 loc) · 8.89 KB

The RAVEN Toolbox 2

The RAVEN (Reconstruction, Analysis and Visualization of Metabolic Networks) Toolbox 2 is a software suite for Matlab that allows for semi-automated reconstruction of genome-scale models (GEMs). It makes use of published models and/or KEGG, MetaCyc databases, coupled with extensive gap-filling and quality control features. The software suite also contains methods for visualizing simulation results and omics data, as well as a range of methods for performing simulations and analyzing the results. The software is a useful tool for system-wide data analysis in a metabolic context and for streamlined reconstruction of metabolic networks based on protein homology.

If you are using RAVEN in any scientific work, please cite: R. Agren, et. al, “The RAVEN Toolbox and Its Use for Generating a Genome-scale Metabolic Model for Penicillium chrysogenum,” PLoS Comput. Biol., vol. 9, no. 3, p. e1002980, Mar. 2013..

A manuscript describing RAVEN Toolbox 2 is currently being prepared. Citation details will therefore be updated in the near future.

Please report any technical issues and bugs here. For other issues, please contact Eduard Kerkhoven.

Releases

RAVEN can be installed via cloning the GitHub repository as per below or by downloading and extracting one of the a zipped release. Please note that the releases do not always represent the most up to date version.

Installation

Required software

  • A functional MATLAB installation (version 2013b or later).

Dependencies

  • libSBML MATLAB API (version 5.15 or higher), which is utilised for importing and exporting GEMs in SBML format. Note: not needed if COBRA Toolbox is installed.
  • At least one solver for linear programming:
    • Preferred: Gurobi Optimizer (version 7.5 or higher), academic license is available here.
    • Alternative/legacy: MOSEK (version 7 only), academic license is available here.
    • If the user has COBRA Toolbox installed, it is possible to use the default COBRA solver (the one which is set by changeCobraSolver).

Instructions

RAVEN Toolbox

Obtain a RAVEN Toolbox in one of the following ways:

  • In Terminal/Command Prompt, navigate to the desired installation directory and run the following Git command:
git clone [email protected]:SysBioChalmers/RAVEN.git
  • Alternatively, download the latest release of RAVEN Toolbox as a ZIP file, and extracted to your favourite directory.

Once extracted, ensure that all other software dependencies (e.g. libSBML, Gurobi) are installed (see above for list, below for instructions. Then, open MATLAB and run the following command:

cd('[location]/RAVEN/installation'))
checkInstallation

where [location] is the directory where you installed RAVEN.

This function checks the functionality for libSBML MATLAB API and solver software. It automatically recognises which solvers are installed and sets the first functional solver as the default RAVEN solver. The default RAVEN solver be changed any time by typing in Matlab:

setRavenSolver('solverName')

Available solver names are gurobi, mosek and cobra.

In Unix-based systems checkInstallation also checks the consistency of external binary programs. If these binaries are broken, they need to be re-compiled from their corresponding source codes. See the documentation for the corresponding software for more details.

libSBML

  1. Download libSBML from the link above and install to your favourite directory.
  2. In MATLAB, run the following command:
addpath('[location]/libSBML-5.x.0-matlab')
savepath

where [location] is where you installed libSBML and 5.x.0 is your libSBML version (5.15.0 or higher).

Gurobi

  1. Download from the link above and install Gurobi to your favourite location.
  2. Make sure you obtained a license following instructions for Windows, Mac or Unix.
  3. To install Gurobi in MATLAB, follow the instructions for Windows, Mac or Unix.
  4. Make sure that MATLAB remembers the Gurobi installation for next time, by running the following command:
savepath

Mosek

  1. Download from the link above and install Mosek to your favourite location.
  2. Make sure you obtained a license following [instructions] (https://docs.mosek.com/8.0/install/installation.html#setting-up-the-license).
  3. To install Mosek in MATLAB, follow instructions. Note: the documentation mentions version 8, but RAVEN only works with version 7 of Mosek.
  4. Make sure that MATLAB remembers the Mosek installation for next time, by running the following command:
savepath

COBRA Toolbox

  1. To gain access to functions from COBRA Toolbox, follow installation instructions provided here.
  2. To use COBRA-specified solvers (e.g. open-source GLPK solver), configure COBRA and RAVEN with the following commands:
changeCobraSolver('glpk')
setRavenSolver('cobra')

Tutorials

Some tutorials highlighting basic RAVEN functionality can be found in the 'tutorial' folder in the installation directory.

Hidden Markov Models (HMMs) for KEGG based reconstruction

Provided are pre-trained Hidden Markov Models (HMMs) for KEGG Orthology (KO) protein sets:

For RAVEN 2.0 and later

For de novo reconstruction of a GEM, the RAVEN function getKEGGModelForOrganism can use HMMs trained on KO protein sets. Provided are HMMs trained on KEGG Release 82.0. CD-HIT was used to obtain non-redundant representative KO protein sets thereby clustering proteins with the defined identity and overlap with the longest protein in the corresponding cluster threshold values. Multisequence alignment with MAFFT and training with HMMER 3.1b2 were then performed. The provided archives contain only pre-trained HMMs.

HMM sets can be downloaded automatically during GEM reconstruction from KEGG (set the dataDir parameter in getKEGGModelForOrganism). Alternatively, download links are provided below. The following HMM sets are available:

dataDir KEGG version Phylogeny Identity (%) Overlap (%)
euk100_kegg82 82.0 eukaryota 100 90
euk90_kegg82 82.0 eukaryota 90 90
euk50_kegg82 82.0 eukaryota 50 90
prok100_kegg82 82.0 prokaryota 100 90
prok90_kegg82 82.0 prokaryota 90 90
prok50_kegg82 82.0 prokaryota 50 90
For RAVEN 1.9.0 or older

HMMs were trained from KO protein sets, based on KEGG Release 58.1. Multisequence alignment was performed with ClustalW2, whereas HMMs were trained with HMMER 2.3. All the associated proteins were used in multisequence alignment and HMMs training. In addition to pre-trained HMMs, the archives also contain multisequence alignment data. The following HMM sets are available:

Dataset KEGG version Phylogeny
eukaryota 58.1 eukaryota
prokaryota 58.1 eukaryota