SOSlib 1.6.0 :: The Tokyo Release Series
(December 17, 2005,
SOSlib is both a programming library (API) and a set of command-line applications for symbolic and numerical analysis of a system of ordinary differential equations (ODEs), derived from a (bio)chemical reaction network encoded in SBML. This release newly features basic sensitivity analysis routines!
We recommed that you use the development version at github. It has lots of improvements over the last official release (1.6.0, 2005). It works with libSBML versions 3.4.1 - 5.x.x. A branch using the latest Sundials exists..
Support via the sourceforge mailing list is still available, but you can also contact us in private ... always happy to help.
SOSlib is both a programming
library and a command-line
application for symbolic and numerical analysis of a system of ordinary
differential equations (ODEs) derived from a chemical reaction network
encoded in the Systems Biology Markup Language (SBML). It is written in ANSI/ISO C
and distributed under the terms of the GNU Lesser General Public License
package employs libSBML's AST (Abstract
Syntax Tree) for formula representation to construct ODE systems, their
Jacobian matrix and other derivatives. CVODES, the sensitivity-enabled
ODE solver in the SUNDIALS package is
used for numerical integration and sensitivity analysis
of stiff and non-stiff ODE systems.
The native API provides fine-grained interfaces to all internal data structures, symbolic operations and numerical routines, enabling the construction of powerful and efficient analytic applications, hybrid solvers or multi-scale models with interfaces to non SBML data sources. Optional modules based on Graphviz and XMGrace allow a quick inspection of a model's structure and dynamics. All functionalities are accessible directly via a command-line application and several example programs.
The capabilities of
SOSlib for sensitivity analysis allow
the implementation of efficient algorithms for parameter
identification. The identification of model parameters and initial
conditions from noisy experimental data is a typical ill-posed inverse
problem and can be formulated in a stable way as a minimization problem
with a data mismatch and a regularization term. In a parameter
identification software based on
SOSlib, the local
(gradient based) search is performed with the interior point optimizer
IpOptusing the capabilities of
SOSlibfor adjoint sensitivity analysis to efficiently compute the gradient of the data mismatch. To stabilize the solutions with respect to noise in the measurements several regularization techniques have been implemented.
SOSlibfor forward sensitivity analysis.
Availability Precompiled binaries of
PIT and example input files are available upon request
from Stefan Müller