Limits...
Reverse engineering of logic-based differential equation models using a mixed-integer dynamic optimization approach.

Henriques D, Rocha M, Saez-Rodriguez J, Banga JR - Bioinformatics (2015)

Bottom Line: This framework aims to simultaneously identify the regulatory structure (represented by binary parameters) and the real-valued parameters that are consistent with the available experimental data, resulting in a logic-based differential equation model.The performance of the method presented here is illustrated with several case studies: a synthetic pathway problem of signaling regulation, a two-component signal transduction pathway in bacterial homeostasis, and a signaling network in liver cancer cells.Supplementary data are available at Bioinformatics online. julio@iim.csic.es or saezrodriguez@ebi.ac.uk.

View Article: PubMed Central - PubMed

Affiliation: Bioprocess Engineering Group, Spanish National Research Council, IIM-CSIC, C/Eduardo Cabello 6, 36208 Vigo, Spain, Centre of Biological Engineering, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal and European Molecular Biology Laboratory, European Bioinformatics Institute, Wellcome Trust Genome Campus, Cambridge, UK Bioprocess Engineering Group, Spanish National Research Council, IIM-CSIC, C/Eduardo Cabello 6, 36208 Vigo, Spain, Centre of Biological Engineering, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal and European Molecular Biology Laboratory, European Bioinformatics Institute, Wellcome Trust Genome Campus, Cambridge, UK.

No MeSH data available.


Related in: MedlinePlus

Case study 1 (synthetic signaling pathway): (a) Histogram of the final objective function achieved by each method (F(x)) across the multiple independent optimization runs. (b) The accuracy of the obtained solutions as a function of the objective function. Each dot describes the results of an independent optimization run
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4565031&req=5

btv314-F3: Case study 1 (synthetic signaling pathway): (a) Histogram of the final objective function achieved by each method (F(x)) across the multiple independent optimization runs. (b) The accuracy of the obtained solutions as a function of the objective function. Each dot describes the results of an independent optimization run

Mentions: Albeit no solver/configuration was able to recover the correct solution in every run, the multi-phase strategy of MPeSS, was the most reliable method, i.e the method which located vicinity of the optimal solution more often. In Figure 3a, the histogram represents the distribution of final values achieved by each method. By combining both problem formulations (relaxed and MINLP), MPeSS is able to arrive to near-globally optimal values in approximately 47% of the runs. However, because MPeSS also has a large tail of poor solutions, the median of the final objective function values is similar to that of eSS and ACOmi. According to the non-parametric Wilcoxon rank-sum test, the three solvers did not show statistically significant differences (see Supplementary Table S2). MITS systematically failed to solve the problem for the considered FE budget. Convergence curves for the tested methods can be found in the Supplementary Materials (Supplementary Figs S2 and S3).Fig. 3.


Reverse engineering of logic-based differential equation models using a mixed-integer dynamic optimization approach.

Henriques D, Rocha M, Saez-Rodriguez J, Banga JR - Bioinformatics (2015)

Case study 1 (synthetic signaling pathway): (a) Histogram of the final objective function achieved by each method (F(x)) across the multiple independent optimization runs. (b) The accuracy of the obtained solutions as a function of the objective function. Each dot describes the results of an independent optimization run
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4565031&req=5

btv314-F3: Case study 1 (synthetic signaling pathway): (a) Histogram of the final objective function achieved by each method (F(x)) across the multiple independent optimization runs. (b) The accuracy of the obtained solutions as a function of the objective function. Each dot describes the results of an independent optimization run
Mentions: Albeit no solver/configuration was able to recover the correct solution in every run, the multi-phase strategy of MPeSS, was the most reliable method, i.e the method which located vicinity of the optimal solution more often. In Figure 3a, the histogram represents the distribution of final values achieved by each method. By combining both problem formulations (relaxed and MINLP), MPeSS is able to arrive to near-globally optimal values in approximately 47% of the runs. However, because MPeSS also has a large tail of poor solutions, the median of the final objective function values is similar to that of eSS and ACOmi. According to the non-parametric Wilcoxon rank-sum test, the three solvers did not show statistically significant differences (see Supplementary Table S2). MITS systematically failed to solve the problem for the considered FE budget. Convergence curves for the tested methods can be found in the Supplementary Materials (Supplementary Figs S2 and S3).Fig. 3.

Bottom Line: This framework aims to simultaneously identify the regulatory structure (represented by binary parameters) and the real-valued parameters that are consistent with the available experimental data, resulting in a logic-based differential equation model.The performance of the method presented here is illustrated with several case studies: a synthetic pathway problem of signaling regulation, a two-component signal transduction pathway in bacterial homeostasis, and a signaling network in liver cancer cells.Supplementary data are available at Bioinformatics online. julio@iim.csic.es or saezrodriguez@ebi.ac.uk.

View Article: PubMed Central - PubMed

Affiliation: Bioprocess Engineering Group, Spanish National Research Council, IIM-CSIC, C/Eduardo Cabello 6, 36208 Vigo, Spain, Centre of Biological Engineering, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal and European Molecular Biology Laboratory, European Bioinformatics Institute, Wellcome Trust Genome Campus, Cambridge, UK Bioprocess Engineering Group, Spanish National Research Council, IIM-CSIC, C/Eduardo Cabello 6, 36208 Vigo, Spain, Centre of Biological Engineering, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal and European Molecular Biology Laboratory, European Bioinformatics Institute, Wellcome Trust Genome Campus, Cambridge, UK.

No MeSH data available.


Related in: MedlinePlus