Warning

This repository has been archived. To follow the latest developments in GRASP refer to https://github.com/dtu-qmcm/GRASP.

GRASP

Introduction

GRASP is a Matlab package to sample thermodynamically feasible kinetic model ensembles. As input it takes an excel file containing the model stoichiometry, a flux distribution, and Gibbs energies.

It also performs Metabolic Control Analysis (MCA) and stability analysis on the resulting ensembles. The resulting ensembles can also be simulated, and individual models can be exported to SBML.

The documentation can be found here.

Table of contents

• Introduction
• Installation
  • Matlab requirements
  • Python requirements (only for the jupyter notebooks in visualization, which are not really part of GRASP)
• Usage
• Create visualizations environment
• Building the documentation
• Running unit tests
• Know issues and limitations

Installation

Matlab requirements

• Matlab's Parallel Computing Toolbox
• Matlab's Bioinformatics Toolbox
• Matlab's Optimization Toobox (to run GRASP in parallel mode)
• Matlab's SimBiology Toolbox (to export models to SBML)

If you don't want to use the Matlab solver to solve the linear programming problems, you can install Gurobi and just specify in the input file (general sheet) that you want to use gurobi instead of linprog.

Python requirements (only for the jupyter notebooks in visualization, which are not really part of GRASP)

• jupyter lab or jupyter notebook
• numpy
• pandas
• scipy
• matplotlib
• altair

Usage

GRASP allows you to:

• build a model ensemble;
• do metabolic control analysis on the ensemble;
• simulate the ensemble.

To do these, you can find example scripts in the examples folder and tutorials in the tutorials folder.

Once you have the results, you can use the jupyter notebooks in the visualization folder to visualize the results.

For more details, see the documentation here.

Create visualizations environment

To use the jupyter notebooks for result visualization (under visualization folder), make sure you first go to the folder visualization, create the conda environment and install the necessary dependencies with:

conda env create -f environment.yml

then activate the conda environment with:

conda activate grasp_viz

and finally install all other dependencies with poetry:

poetry install

Building the documentation

The documentation is done in sphinx, to build it locally start by going to the folder docs and creating a conda environment with:

conda env create -f environment.yml

then activate the conda environment with:

conda activate grasp_docs

and finally install all other dependencies with poetry:

poetry install

Once all the dependencies are installed, you can just run make html while on the docs folder and that will create a folder _build/html where you can find the html documentation.

Running unit tests

To run the unit tests go to matlab_code/tests and run the script run_unit_tests.m.

Know issues and limitations

• massAction mechanisms only work for one substrate and one product;
• if the uncertainty in the flux is such that it can be both positive and negative, it is possible that GRASP samples Gibbs energies incompatible with the reference flux. This generates an error;
• when using altair in the jupyter notebooks you might get the error <VegaLite 2 object> when trying to plot something.
  • you should add alt.renderers.enable('default') after importing altair.