CoChat

‘Glocal’ Robustness Analysis and Model Discrimination for Circadian Oscillators

Open access
Hafner, Marc, Koeppl, Heinz, Hasler, Martin, Wagner, Andreas · 2009 · PLoS Computational Biology

The DOI matches its Crossref record; 66 of the 68 references it cites were found in a registry or an index.

Abstract

To characterize the behavior and robustness of cellular circuits with many unknown parameters is a major challenge for systems biology. Its difficulty rises exponentially with the number of circuit components. We here propose a novel analysis method to meet this challenge. Our method identifies the region of a high-dimensional parameter space where a circuit displays an experimentally observed behavior. It does so via a Monte Carlo approach guided by principal component analysis, in order to allow efficient sampling of this space. This 'global' analysis is then supplemented by a 'local' analysis, in which circuit robustness is determined for each of the thousands of parameter sets sampled in the global analysis. We apply this method to two prominent, recent models of the cyanobacterial circadian oscillator, an autocatalytic model, and a model centered on consecutive phosphorylation at two sites of the KaiC protein, a key circadian regulator. For these models, we find that the two-sites architecture is much more robust than the autocatalytic one, both globally and locally, based on five different quantifiers of robustness, including robustness to parameter perturbations and to molecular noise. Our 'glocal' combination of global and local analyses can also identify key causes of high or low robustness. In doing so, our approach helps to unravel the architectural origin of robust circuit behavior. Complementarily, identifying fragile aspects of system behavior can aid in designing perturbation experiments that may discriminate between competing mechanisms and different parameter sets.

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 · 10.1371/journal.pcbi.1000534

Cites

66 of 66 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Glocal robustness analysis flow for a hypothetical two dimensional parameter space. (A) A model structure and systemic properties serve as inputs for the global step of the analysis. This global analysis is composed of (B) and (C) and yields viable parameter vectors for the model in addition to the normalized viable volume . Different local perturbations are applied to these parameter vectors (D) in order to quantify their local robustness . (B) Monte Carlo sampling to define viable parameter ranges. The first sampling step uses Gaussian random sampling with independently and identically distributed random variables. Some of the tested parameter vectors (gray circles) are viable (blue circles). For subsequent iterative steps, sampling occurs according to the covariance matrix of viable parameters estimated in previous steps. (C) Monte Carlo integration. To estimate the volume in which viable parameter sets occur, we define a hyperbox (red rectangle in left panel) that contains all the viable parameters of the last iteration. We then sample uniformly parameter vectors from this box (right panel, gray circles) and estimate the fraction of viable parameter vectors (right panel, black circles). (D) Local analyses are performed on all viable parameter vectors and help identify correlations between parameter vectors or their components, and robustness values (color intensity) to provide regions of high robustness in the parameters space; two different local robustness quantifiers (left-red, and right-green).
Figure 1 Glocal robustness analysis flow for a hypothetical two dimensional parameter space. (A) A model structure and systemic properties serve as inputs for the global step of the analysis. This global analysis is composed of (B) and (C) and yields viable parameter vectors for the model in addition to the normalized viable volume . Different local perturbations are applied to these parameter vectors (D) in order to quantify their local robustness . (B) Monte Carlo sampling to define viable parameter ranges. The first sampling step uses Gaussian random sampling with independently and identically distributed random variables. Some of the tested parameter vectors (gray circles) are viable (blue circles). For subsequent iterative steps, sampling occurs according to the covariance matrix of viable parameters estimated in previous steps. (C) Monte Carlo integration. To estimate the volume in which viable parameter sets occur, we define a hyperbox (red rectangle in left panel) that contains all the viable parameters of the last iteration. We then sample uniformly parameter vectors from this box (right panel, gray circles) and estimate the fraction of viable parameter vectors (right panel, black circles). (D) Local analyses are performed on all viable parameter vectors and help identify correlations between parameter vectors or their components, and robustness values (color intensity) to provide regions of high robustness in the parameters space; two different local robustness quantifiers (left-red, and right-green).CC BY · PubMed Central
Figure 2 Two models of the cyanobacterial circadian cycle. (A) Autocatalytic model from Mehra et al. [31]. ‘C*’ stands for phosphorylated KaiC. The cycle proceeds clockwise, starting from the upper left. The sum of concentrations of the KaiC*-containing complexes (underlined) form the output of the model. The red arrow denotes the autocatalytic effect of KaiAC* on its synthesis. (B) Two phosphorylation sites model from Rust et al. [32]. There are three possible phosphorylated states for KaiC: , and . The sum of concentrations of phosphorylated KaiC molecules (underlined) is the output of the system. KaiA catalyzes phosphorylation reactions (solid blue arrows) and inhibits some dephosphorylation reactions (dashed blue bars). (complexed with KaiB, not explicitly modeled) inhibits the action of KaiA (red bar).
Figure 2 Two models of the cyanobacterial circadian cycle. (A) Autocatalytic model from Mehra et al. [31]. ‘C*’ stands for phosphorylated KaiC. The cycle proceeds clockwise, starting from the upper left. The sum of concentrations of the KaiC*-containing complexes (underlined) form the output of the model. The red arrow denotes the autocatalytic effect of KaiAC* on its synthesis. (B) Two phosphorylation sites model from Rust et al. [32]. There are three possible phosphorylated states for KaiC: , and . The sum of concentrations of phosphorylated KaiC molecules (underlined) is the output of the system. KaiA catalyzes phosphorylation reactions (solid blue arrows) and inhibits some dephosphorylation reactions (dashed blue bars). (complexed with KaiB, not explicitly modeled) inhibits the action of KaiA (red bar).CC BY · PubMed Central
Figure 3 Results of the global robustness analyses for both models. (A) The two-sites model (right) has significantly greater nomalized viable volume than the autocatalytic model (left). Error bars () correspond to standard deviations over five independent estimates. (B) Standard deviations along the principal axes of viable parameters for the autocatalytic model and the two-sites model. Note the logarithmic scale. The autocatalytic model has a strongly constrained axis (arrow); amounts of variation along the other axes are overall smaller for the autocatalytic model. (C) Projection of the viable vectors of the autocatalytic model after the MC integration on the plane . These two parameters are strongly correlated resulting in the lowest standard deviation for the autocatalytic model (B).
Figure 3 Results of the global robustness analyses for both models. (A) The two-sites model (right) has significantly greater nomalized viable volume than the autocatalytic model (left). Error bars () correspond to standard deviations over five independent estimates. (B) Standard deviations along the principal axes of viable parameters for the autocatalytic model and the two-sites model. Note the logarithmic scale. The autocatalytic model has a strongly constrained axis (arrow); amounts of variation along the other axes are overall smaller for the autocatalytic model. (C) Projection of the viable vectors of the autocatalytic model after the MC integration on the plane . These two parameters are strongly correlated resulting in the lowest standard deviation for the autocatalytic model (B).CC BY · PubMed Central
Figure 4 The two-sites model (blue) has greater local robustness than the autocatalytic model (red). Shown are the distributions of (A) robustness to local parameter perturbations , (B) robustness to total concentration perturbations , (C) robustness to molecular noise , (D) attraction of the cycle , and (E) sensitivity of the period . In (F) median values are shown with their associated standard deviation (error bars) for both models and all five quantifiers. Black dots indicate local robustness values for the previously published parameter vector's [31],[32].
Figure 4 The two-sites model (blue) has greater local robustness than the autocatalytic model (red). Shown are the distributions of (A) robustness to local parameter perturbations , (B) robustness to total concentration perturbations , (C) robustness to molecular noise , (D) attraction of the cycle , and (E) sensitivity of the period . In (F) median values are shown with their associated standard deviation (error bars) for both models and all five quantifiers. Black dots indicate local robustness values for the previously published parameter vector's [31],[32].CC BY · PubMed Central
Figure 5 Distribution of the average local robustness for the two models. For each viable parameter vector , the figure shows its distance (horizontal axis) from the viable parameter vector with the highest average local robustness plotted against the of (vertical axis); autocatalytic model (red) and two-sites model (blue). Large circles correspond to the two parameter vectors with the highest for each model, and squares correspond to published parameter vectors [31],[32]. The greater the distance of to the most robust parameter vector, the lower its . This negative association is stronger for the autocatalytic model.
Figure 5 Distribution of the average local robustness for the two models. For each viable parameter vector , the figure shows its distance (horizontal axis) from the viable parameter vector with the highest average local robustness plotted against the of (vertical axis); autocatalytic model (red) and two-sites model (blue). Large circles correspond to the two parameter vectors with the highest for each model, and squares correspond to published parameter vectors [31],[32]. The greater the distance of to the most robust parameter vector, the lower its . This negative association is stronger for the autocatalytic model.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

Cited by

See all 63

Metadata

Authors
Hafner, M., Koeppl, H., Hasler, M., & Wagner, A.
Affiliations
  • SIB Swiss Institute of Bioinformatics (CH)
  • École Polytechnique Fédérale de Lausanne (CH)
Corresponding author
Wagner, Andreas
Venue
PLoS Computational Biology
Published
2009-10-16
Volume
5
Issue
10
Pages
e1000534
ISSN
  • 1553-734X (ISSN-L)
  • 1553-7358
OpenAlex
W2049648266
Cited by
63
References
68
Citations by year
2026: 1 · 2022: 2 · 2021: 1 · 2020: 4 · 2019: 2 · 2018: 3 · 2017: 3 · 2016: 3 · 2015: 5 · 2014: 6 · 2013: 9 · 2012: 8
FWCI
3.59
Type
journal article
Language
en
Open access
Locations
  • PLoS Computational Biology (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Zurich Open Repository and Archive (University of Zurich) (acceptedVersion, cc-by)
  • TUbilio (Technical University of Darmstadt) (acceptedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Figshare (submittedVersion, cc-by)
  • Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion)
  • Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion, cc-by-nc-nd)
  • TUbilio (Technical University of Darmstadt) (submittedVersion, cc-by)
  • Technischen Universität Darmstadt (submittedVersion, cc-by)
  • Technischen Universität Darmstadt (submittedVersion)
  • Zurich Open Repository and Archive (University of Zurich) (submittedVersion, cc-by)
  • Technischen Universität Darmstadt (cc-by)
  • Zurich Open Repository and Archive (University of Zurich)
Funders
  • National Science Foundation (200020)
  • Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung (315200, 119697, 200020, 117975, 116814)
  • SystemsX.ch
Topics
  • Life Sciences › Neuroscience › Endocrine and Autonomic Systems › Circadian rhythm and melatonin
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene Regulatory Network Analysis
  • Physical Sciences › Mathematics › Modeling and Simulation › Mathematical Biology Tumor Growth
Keywords
circadian clock, cyanobacterial circadian clock, KaiABC, KaiC phosphorylation, autocatalytic model, robustness analysis, parameter space exploration, principal component analysis, Monte Carlo sampling, molecular noise, model discrimination, circuit architecture, parameter uncertainty
MeSH terms
Circadian Rhythm, Models, Theoretical

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

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.