Multiscale Systems Biology Collaboration

In silico models

78 M models from BiGG

JSON data files for 78 Flux Balance Analysis models (LP problems) from the BiGG database. Python code to apply Double MINOS and Quad MINOS to these models is here: quadLPpy.zip.

83 FBA models

MPS data files for 83 Flux Balance Analysis models (LP problems) corresponding to 83 Matlab structures in a file “all_models.mat” generated by running the following notebook written by Ali Ebrahim, UC San Diego: load_models.ipynb. The MPS files were generated by these Matlab scripts: dumpMPS2.m and dumpMPS2all.m.

quadLP data

MPS data files for 14 multiscale LP models used in this paper: D. Ma, L. Yang, R. M. T. Fleming, I. Thiele, B. O. Palsson and M. A. Saunders, Reliable and efficient solution of genome-scale models of Metabolism and macromolecular Expression, Sci. Rep. 7, Article number: 40863 (2017); doi:10.1038/srep40863.

openCOBRA toolbox

COnstraints Based Reconstruction and Analysis toolbox for MATLAB.

BiGG Database, Systems Biology Research Group, University of California, San Diego

Molecular Systems Physiology Group, University of Luxembourg (Dr Ines Thiele)

Algorithm NCL for nonlinearly constrained optimization

Optimal taxation policy

AMPL scripts, documentation, and results for solving Optimal Taxation problems with many nonlinear constraints that don't satisfy the LICQ. Algorithm NCL generates a sequence of nonlinearly constrained augmented Lagrangian subproblems analogous to the bound constrained subproblems generated by LANCELOT. The taxation problem and Algorithm NCL are modeled in AMPL and solved by the interior method IPOPT.

The files here illustrate how Algorithm NCL can be implemented in a high-level language when first and derivatives of the problem objective and constraints are available via that language (as in AMPL and GAMS).