Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 38 of the 38 references it cites were found in a registry or an index.

A cooperative strategy for parameter estimation in large scale systems biology models

Open access
Villaverde, Alejandro F, Egea, Jose A, Banga, Julio R · 2012 · BMC Systems Biology

Abstract

Background: Mathematical models play a key role in systems biology: they summarize the currently available knowledge in a way that allows to make experimentally verifiable predictions. Model calibration consists of finding the parameters that give the best fit to a set of experimental data, which entails minimizing a cost function that measures the goodness of this fit. Most mathematical models in systems biology present three characteristics which make this problem very difficult to solve: they are highly non-linear, they have a large number of parameters to be estimated, and the information content of the available experimental data is frequently scarce. Hence, there is a need for global optimization methods capable of solving this problem efficiently. Results: A new approach for parameter estimation of large scale models, called Cooperative Enhanced Scatter Search (CeSS), is presented. Its key feature is the cooperation between different programs (“threads”) that run in parallel in different processors. Each thread implements a state of the art metaheuristic, the enhanced Scatter Search algorithm (eSS). Cooperation, meaning information sharing between threads, modifies the systemic properties of the algorithm and allows to speed up performance. Two parameter estimation problems involving models related with the central carbon metabolism of E. coli which include different regulatory levels (metabolic and transcriptional) are used as case studies. The performance and capabilities of the method are also evaluated using benchmark problems of large-scale global optimization, with excellent results. Conclusions: The cooperative CeSS strategy is a general purpose technique that can be applied to any model calibration problem. Its capability has been demonstrated by calibrating two large-scale models of different characteristics, improving the performance of previously existing methods in both cases. The cooperative metaheuristic presented here can be easily extended to incorporate other global and local search solvers and specific structural information for particular classes of problems.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1186/1752-0509-6-75

Cites

38 of 38 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 eSS algorithm. Enhanced scatter search scheme.
Figure 1 eSS algorithm. Enhanced scatter search scheme.CC BY · PubMed Central
Figure 2 Cooperative algorithm. Scheme of the cooperative algorithm; η threads exchange information at fixed intervals with periodicity τ, as defined in eq. (8).
Figure 2 Cooperative algorithm. Scheme of the cooperative algorithm; η threads exchange information at fixed intervals with periodicity τ, as defined in eq. (8).CC BY · PubMed Central
Figure 3 Model 1 - Histograms of multistart local searches. Objective function values achieved by local searches starting from different points with the FMINSEARCH (Figure3.a), DHC (Figure3.b), and FMINCON (Figure3.c) methods. Calculations were carried out in a single processor and the computation time was 10 days for each method. The best values obtained were: 5.367·105(FMINSEARCH), 2.938·101(DHC), and 1.459·103(FMINCON).
Figure 3 Model 1 - Histograms of multistart local searches. Objective function values achieved by local searches starting from different points with the FMINSEARCH (Figure3.a), DHC (Figure3.b), and FMINCON (Figure3.c) methods. Calculations were carried out in a single processor and the computation time was 10 days for each method. The best values obtained were: 5.367·105(FMINSEARCH), 2.938·101(DHC), and 1.459·103(FMINCON).CC BY · PubMed Central
Figure 4 Model 1 - Differential Evolution vs. non-cooperative enhanced Scatter Search. Convergence curves of the DE and eSS methods. For each method 10 threads were launched; the figure shows at every instant the best value found by the threads.
Figure 4 Model 1 - Differential Evolution vs. non-cooperative enhanced Scatter Search. Convergence curves of the DE and eSS methods. For each method 10 threads were launched; the figure shows at every instant the best value found by the threads.CC BY · PubMed Central
Figure 5 Model 1 - Convergence curves, CeSS vs. eSS. The performance of 10 individual, non cooperative threads (in black) is compared with η = 10 cooperative threads that exchange information with τ = 2.53 (in green). The figure shows the logarithm of the objective function value plotted against the computation time. For each method, only the best value found by its 10 threads is shown at every instant.
Figure 5 Model 1 - Convergence curves, CeSS vs. eSS. The performance of 10 individual, non cooperative threads (in black) is compared with η = 10 cooperative threads that exchange information with τ = 2.53 (in green). The figure shows the logarithm of the objective function value plotted against the computation time. For each method, only the best value found by its 10 threads is shown at every instant.CC BY · PubMed Central
Figure 6 Model 1 - relative errors in the model predictions. Figure6.a plots the values of the measured vs. predicted states of the model. Note that the plot range is from 10−5to 10−4; 10 of the 1150 points have values less than 10−6and are not shown. Figure6.b shows an histogram of the residuals, which are calculated as |(x~−x)/x|·100, where the predicted values are x~. In both figures the results are plotted in a logarithmic scale.
Figure 6 Model 1 - relative errors in the model predictions. Figure6.a plots the values of the measured vs. predicted states of the model. Note that the plot range is from 10−5to 10−4; 10 of the 1150 points have values less than 10−6and are not shown. Figure6.b shows an histogram of the residuals, which are calculated as |(x~−x)/x|·100, where the predicted values are x~. In both figures the results are plotted in a logarithmic scale.CC BY · PubMed Central
Figure 7 Model 1 - relative errors in the estimated parameters.Figure7.a plots the values of the nominal vs. estimated parameters. Figure7.b shows an histogram of the residuals, which are calculated as |(p~−p)/p|·100, where the estimated parameters are p~. In both figures the results are plotted in a logarithmic scale.
Figure 7 Model 1 - relative errors in the estimated parameters.Figure7.a plots the values of the nominal vs. estimated parameters. Figure7.b shows an histogram of the residuals, which are calculated as |(p~−p)/p|·100, where the estimated parameters are p~. In both figures the results are plotted in a logarithmic scale.CC BY · PubMed Central
Figure 8 Model 2 - Histograms of multistart local searches. Objective function values achieved by local searches starting from different points with the FMINSEARCH (Figure8.a), DHC (Figure8.b), and N2FB (Figure8.c) methods. Calculations were carried out in a single processor; the computation time was 240 hours for each method. The best values obtained were: 3.046·103(FMINSEARCH), 7.601·102(DHC), and 1.3863·103(N2FB). Note that the x axis is logarithmic in order to represent the large dispersion of the results.
Figure 8 Model 2 - Histograms of multistart local searches. Objective function values achieved by local searches starting from different points with the FMINSEARCH (Figure8.a), DHC (Figure8.b), and N2FB (Figure8.c) methods. Calculations were carried out in a single processor; the computation time was 240 hours for each method. The best values obtained were: 3.046·103(FMINSEARCH), 7.601·102(DHC), and 1.3863·103(N2FB). Note that the x axis is logarithmic in order to represent the large dispersion of the results.CC BY · PubMed Central
Figure 9 Model 2 - Differential Evolution vs. non-cooperative enhanced Scatter Search. Convergence curves of the DE and eSS methods. For each method 10 threads were launched; the figure shows at every instant the best value found by the threads.
Figure 9 Model 2 - Differential Evolution vs. non-cooperative enhanced Scatter Search. Convergence curves of the DE and eSS methods. For each method 10 threads were launched; the figure shows at every instant the best value found by the threads.CC BY · PubMed Central
Figure 10 Model 2 - Convergence curves, CeSS vs. eSS. The logarithm of the objective function value is plotted against the computation time. The performance of 10 individual, non cooperative threads is shown in black. It is compared with η = 10 cooperative threads that exchange information with τ = 2.15 (green line). For each method, only the best value found by its 10 threads is shown at every instant
Figure 10 Model 2 - Convergence curves, CeSS vs. eSS. The logarithm of the objective function value is plotted against the computation time. The performance of 10 individual, non cooperative threads is shown in black. It is compared with η = 10 cooperative threads that exchange information with τ = 2.15 (green line). For each method, only the best value found by its 10 threads is shown at every instantCC BY · PubMed Central
Figure 11 Model 2 - Data vs. model predictions. Dynamics of the 47 model states. Data points used for calibration are shown as blue dots; the calibrated model predictions are shown as red lines. The plots show a near perfect match between data and predictions.
Figure 11 Model 2 - Data vs. model predictions. Dynamics of the 47 model states. Data points used for calibration are shown as blue dots; the calibrated model predictions are shown as red lines. The plots show a near perfect match between data and predictions.CC BY · PubMed Central
Figure 12 Model 2 - relative errors in the model predictions.Figure12.a plots the values of the measured vs. predicted states of the model. For clarity, values corresponding to very small concentrations (<10−6) are not shown. Figure12.b shows an histogram of the residuals, which are calculated as |(x~−x)/x|·100, where the predicted values are x~.
Figure 12 Model 2 - relative errors in the model predictions.Figure12.a plots the values of the measured vs. predicted states of the model. For clarity, values corresponding to very small concentrations (<10−6) are not shown. Figure12.b shows an histogram of the residuals, which are calculated as |(x~−x)/x|·100, where the predicted values are x~.CC BY · PubMed Central
Figure 13 Model 2 - relative errors in the estimated parameters. Figure13.a plots the values of the nominal vs. estimated parameters. Figure13.b shows an histogram of the residuals, which are calculated as |(p~−p)/p|·100, where the estimated parameters are p~. In both figures the results are plotted in a logarithmic scale. The 11 parameters with values <10−50are not shown.
Figure 13 Model 2 - relative errors in the estimated parameters. Figure13.a plots the values of the nominal vs. estimated parameters. Figure13.b shows an histogram of the residuals, which are calculated as |(p~−p)/p|·100, where the estimated parameters are p~. In both figures the results are plotted in a logarithmic scale. The 11 parameters with values <10−50are not shown.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 38 references

Research works printed but not linked: 2

  1. De La Maza M, Yuret D: Dynamic hill climbing. AI EXPERT 1994, 9: 26-26.
  2. IEEE World Congress on Computational Intelligence (CEC@WCCI-2012): Competition on Large Scale Global Optimization. 2012.http://staff.ustc.edu.cn/~ketang/cec2012/lsgo_competition.htm []

Cited by

See all 56

Metadata

Authors
Villaverde, A. F., Egea, J. A., & Banga, J. R.
Affiliations
Consejo Superior de Investigaciones Científicas (ES)
Corresponding author
Banga, Julio R
Venue
BMC Systems Biology
Published
2012-06-22
Volume
6
Issue
1
Pages
75-75
ISSN
1752-0509 (ISSN-L)
OpenAlex
W2090517830
Cited by
56
References
38
Citations by year
2024: 4 · 2023: 3 · 2022: 2 · 2021: 1 · 2020: 3 · 2019: 6 · 2018: 6 · 2017: 5 · 2016: 2 · 2015: 11 · 2014: 6 · 2013: 7
FWCI
7.56
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • BMC Systems Biology (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DIGITAL.CSIC (Spanish National Research Council (CSIC)) (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Funders
  • European Commission (289434)
  • Ministerio de Economía y Competitividad (DPI2011-28112-C04-04, DPI2011-28112-C04-03, C04-04)
  • FP7 Food, Agriculture and Fisheries, Biotechnology
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Metaheuristic Optimization Algorithms Research
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene Regulatory Network Analysis
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Microbial Metabolic Engineering and Bioproduction
Keywords
parameter estimation, model calibration, systems biology models, cooperative metaheuristics, global optimization, parallel threads, information sharing, central carbon metabolism, Escherichia coli, transcriptional regulation, metabolic regulation, large-scale optimization
MeSH terms
Algorithms, Calibration, Carbon, Escherichia coli, Models, Biological, Time Factors, Transcription, Genetic, Gene Expression Regulation, Bacterial, Systems Biology

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: A cooperative strategy for parameter estimation in large scale systems biology models. https://app.cochat.ai/papers/10.1186/1752-0509-6-75
The paper
Villaverde, A. F., Egea, J. A., & Banga, J. R. (2012). A cooperative strategy for parameter estimation in large scale systems biology models. BMC Systems Biology. https://doi.org/10.1186/1752-0509-6-75

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.