Latest stable release: 1.1.0, released 23 June 2016 (see box at right →)
MOCCASIN can read certain forms of ODE-based models written in MATLAB and translate them into SBML format. It does not use or need MATLAB itself; it is written in Python, and is available under LGPL terms. It offers both a GUI and a command-line interface.
Please cite the paper if you use MOCCASIN
Article citations are crucial to our ability to justify continued development of MOCCASIN, as well as our academic careers. If you use MOCCASIN and you publish papers about your software, we ask that you please cite the paper:
Harold F. Gómez, Michael Hucka, Sarah M. Keating, German Nudelman, Dagmar Iber and Stuart C. Sealfon. MOCCASIN: converting MATLAB ODE models to SBML. Bioinformatics (2016), 32(12): 1905-1906. (Freely available directly from Bioinformatics.)
MOCCASIN release DOIs
We provide Zenodo DOIs for specific releases, so that you can cite the exact version of the software that you use:
- MOCCASIN release 1.1.0 ⇒ 10.5281/zenodo.56374
More about MOCCASIN
MOCCASIN's goal is to to use a combination of heuristics and user assistance to help researchers export models written as ordinary MATLAB and Octave scripts. It uses an algorithm developed by Fages, Gay and Soliman described in the paper titled Inferring reaction systems from ordinary differential equations. A free technical report explaining the algorithm is available from INRIA.
To parse MATLAB and produce input to the reaction-inference algorithm, MOCCASIN uses a custom MATLAB parser written using PyParsing, combined with heuristics and a variety of post-processing operations to interpret the MATLAB contents.
Currently, MOCCASIN is limited to MATLAB inputs in which a model is contained in a single file. The file must set up a system of differential equations as a function defined in the file, and make a call to one of the MATLAB odeNN family of solvers (e.g.,
ode15s, etc.). While this may seem limiting, it is in fact a common form of MATLAB models used in systems biology.
Current and future work
MOCCASIN can be significantly improved, in particular in its ability to translate more MATLAB constructs. We welcome your help! If you are interested, please get in touch with the developers so that we can coordinate better.
This work is made possible thanks in part to funding from the Icahn School of Medicine at Mount Sinai, provided as part of the NIH-funded project Modeling Immunity for Biodefense (contract number HHSN266200500021C, principal Investigator: Stuart Sealfon), and in part to funding from the School of Medicine at Boston University, provided as part of the NIH-funded project Modeling Immunity for Biodefense (contract number HHSN272201000053C, principal Investigators: Thomas B. Kepler and Garnett H. Kelsoe).
We also acknowledge the contributions made by Dr. Dagmar Iber from the Department of Biosystems Science and Engineering (D-BSSE), and Dr. Bernd Rinn from the Scientific IT Services (SIS) division from ETH Zurich.