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

The DOI matches its Crossref record; 87 of the 88 references it cites were found in a registry or an index, and 1 are not checked yet.

Data-driven reverse engineering of signaling pathways using ensembles of dynamic models

Open access
Henriques, David, Villaverde, Alejandro F., Rocha, Miguel, Saez-Rodriguez, Julio, Banga, Julio R. · 2017 · PLOS Computational Biology

Abstract

Despite significant efforts and remarkable progress, the inference of signaling networks from experimental data remains very challenging. The problem is particularly difficult when the objective is to obtain a dynamic model capable of predicting the effect of novel perturbations not considered during model training. The problem is ill-posed due to the nonlinear nature of these systems, the fact that only a fraction of the involved proteins and their post-translational modifications can be measured, and limitations on the technologies used for growing cells in vitro, perturbing them, and measuring their variations. As a consequence, there is a pervasive lack of identifiability. To overcome these issues, we present a methodology called SELDOM (enSEmbLe of Dynamic lOgic-based Models), which builds an ensemble of logic-based dynamic models, trains them to experimental data, and combines their individual simulations into an ensemble prediction. It also includes a model reduction step to prune spurious interactions and mitigate overfitting. SELDOM is a data-driven method, in the sense that it does not require any prior knowledge of the system: the interaction networks that act as scaffolds for the dynamic models are inferred from data using mutual information. We have tested SELDOM on a number of experimental and in silico signal transduction case-studies, including the recent HPN-DREAM breast cancer challenge. We found that its performance is highly competitive compared to state-of-the-art methods for the purpose of recovering network topology. More importantly, the utility of SELDOM goes beyond basic network inference (i.e. uncovering static interaction networks): it builds dynamic (based on ordinary differential equation) models, which can be used for mechanistic interpretations and reliable dynamic predictions in new experimental conditions (i.e. not used in the training). For this task, SELDOM's ensemble prediction is not only consistently better than predictions from individual models, but also often outperforms the state of the art represented by the methods used in the HPN-DREAM challenge.

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.1005379

Cites

87 of 88 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig 1 SELDOM workflow. The experimental data is used to build an adjacency (a dense DDN) matrix based on the mutual information of all pairs of variables. Through a simple sampling scheme, and limiting the maximum in-degree for each node, a set of more sparse DDNs are generated. Each individual DDN is then used as a scaffold for independent model training and model reduction problems. The resulting models are used to form an ensemble which is able to produce predictions for state trajectories under untested experimental conditions.
Fig 1 SELDOM workflow. The experimental data is used to build an adjacency (a dense DDN) matrix based on the mutual information of all pairs of variables. Through a simple sampling scheme, and limiting the maximum in-degree for each node, a set of more sparse DDNs are generated. Each individual DDN is then used as a scaffold for independent model training and model reduction problems. The resulting models are used to form an ensemble which is able to produce predictions for state trajectories under untested experimental conditions.CC BY · PubMed Central
Fig 2 MAPK signaling network. The model by Huang et al. [67] was used to generate pseudo-experimental data for two sub-problems. The first (MAPKp) partially observed (MAPK-PP, MAPKK-PP and MAPKKK), and the second fully observed MAPKf.
Fig 2 MAPK signaling network. The model by Huang et al. [67] was used to generate pseudo-experimental data for two sub-problems. The first (MAPKp) partially observed (MAPK-PP, MAPKK-PP and MAPKKK), and the second fully observed MAPKf.CC BY · PubMed Central
Fig 3 Time course predictions for case study 1b (MAPKf). The median in red is surrounded by the predicted non-symmetric 20%,60% and 95% confidence intervals.
Fig 3 Time course predictions for case study 1b (MAPKf). The median in red is surrounded by the predicted non-symmetric 20%,60% and 95% confidence intervals.CC BY · PubMed Central
Fig 4 Relationship between training and prediction RMSE for case study 3 (DREAMiS). The prediction RMSE is plotted here against the training RMSE for each individual model (blue) and the ensemble (red). RMSE scores for the top 3 performing teams in the HPN-DREAM in silico time-course prediction sub-challenge (Team34, Team8 and Team10) are also shown as colored lines.
Fig 4 Relationship between training and prediction RMSE for case study 3 (DREAMiS). The prediction RMSE is plotted here against the training RMSE for each individual model (blue) and the ensemble (red). RMSE scores for the top 3 performing teams in the HPN-DREAM in silico time-course prediction sub-challenge (Team34, Team8 and Team10) are also shown as colored lines.CC BY · PubMed Central
Fig 5 Heatmap with RMSE scores for different methods and case studies. The color scheme represents RMSE scores normalized by case-study in order to emphasize differences between methods. The color scale moves from green (low RMSE) to blue (high RMSE). The numeric values of the RMSE scores for each method/case-study are also provided in each corresponding cell. SELDOM B and SELDOM C were clearly the most robust strategies doing very well in all problems.
Fig 5 Heatmap with RMSE scores for different methods and case studies. The color scheme represents RMSE scores normalized by case-study in order to emphasize differences between methods. The color scale moves from green (low RMSE) to blue (high RMSE). The numeric values of the RMSE scores for each method/case-study are also provided in each corresponding cell. SELDOM B and SELDOM C were clearly the most robust strategies doing very well in all problems.CC BY · PubMed Central
Fig 6 Ensemble predictive skill depending on ensemble size (case study DREAMiS). This curve was computed by bootstrapping multiple nℳ models from the available 100 models, i.e. we sampled multiple realizations of the individual predictions for the same ensemble size and computed the average value. These curves converge asymptotically and show that the chosen ensemble size parameter is adequate. Equivalent predictions could have been obtained with smaller ensemble sizes.
Fig 6 Ensemble predictive skill depending on ensemble size (case study DREAMiS). This curve was computed by bootstrapping multiple nℳ models from the available 100 models, i.e. we sampled multiple realizations of the individual predictions for the same ensemble size and computed the average value. These curves converge asymptotically and show that the chosen ensemble size parameter is adequate. Equivalent predictions could have been obtained with smaller ensemble sizes.CC BY · PubMed Central
Fig 7 Heatmap with AUPR scores for different methods and case studies. The color scheme represents AUPR scores normalized by case-study in order to emphasize differences between methods. The color scale moves from red (low AUPR) to yellow (high AUPR). The numeric values of the AUPR scores for each method/case-study are also provided in each corresponding cell. The sparsest version of SELDOM (A) did consistently well in all the case studies. SELDOM B and C did an average job with MAPKf but provided good solutions for all other case-studies. The comparisons are only provided for in silico problems with known solution. Additionally, the solution for the top performing teams in the DREAM challenge is only available for DREAMiS.
Fig 7 Heatmap with AUPR scores for different methods and case studies. The color scheme represents AUPR scores normalized by case-study in order to emphasize differences between methods. The color scale moves from red (low AUPR) to yellow (high AUPR). The numeric values of the AUPR scores for each method/case-study are also provided in each corresponding cell. The sparsest version of SELDOM (A) did consistently well in all the case studies. SELDOM B and C did an average job with MAPKf but provided good solutions for all other case-studies. The comparisons are only provided for in silico problems with known solution. Additionally, the solution for the top performing teams in the DREAM challenge is only available for DREAMiS.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 83 references

Research works printed but not linked: 3

  1. TG Dietterich, 2000, Multiple Classifier Systems: First International Workshop, 1
  2. L Breiman, 1998, Arcing classifiers, Ann Stat, 26, 801
  3. M De La Maza, 1994, Dynamic hill climbing, AI expert, 9, 26

Cited by

See all 46

Metadata

Authors
Henriques, D., Villaverde, A. F., Rocha, M., Saez-Rodriguez, J., & Banga, J. R.
Affiliations
  • Consejo Superior de Investigaciones Científicas (ES)
  • University of Minho (PT)
Corresponding author
Banga, Julio R.
Venue
PLOS Computational Biology
Published
2017-02-06
Volume
13
Issue
2
Pages
e1005379
ISSN
  • 1553-734X (ISSN-L)
  • 1553-7358
OpenAlex
W2586633795
Cited by
44
References
83
Citations by year
2025: 1 · 2024: 5 · 2023: 2 · 2022: 5 · 2021: 4 · 2020: 8 · 2019: 8 · 2018: 7 · 2017: 4
FWCI
3.02
Type
journal article
Language
en
Open access
Locations
  • PLoS Computational Biology (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DIGITAL.CSIC (Spanish National Research Council (CSIC)) (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Figshare (submittedVersion, cc-by)
  • RWTH Publications (RWTH Aachen) (submittedVersion, other-oa)
  • Portuguese National Funding Agency for Science, Research and Technology (RCAAP Project by FCT) (submittedVersion, other-oa)
  • RWTH Publications (RWTH Aachen)
Funders
  • Galician government (Xunta de Galiza) I2C (ED481B2014/133-0)
  • EU FP7 (ITN Grant number 289384)
  • Spanish MINECO (DPI2014-55276-C5-2-R)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene Regulatory Network Analysis
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Microbial Metabolic Engineering and Bioproduction
Keywords
signaling pathway inference, ensemble forecasting, model reduction, network identifiability, signal transduction, ordinary differential equation model
MeSH terms
Algorithms, Animals, Computer Simulation, Humans, Models, Biological, Signal Transduction, Proteome, Metabolic Engineering, Metabolic Flux Analysis

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: Data-driven reverse engineering of signaling pathways using ensembles of dynamic models. https://app.cochat.ai/papers/10.1371/journal.pcbi.1005379
The paper
Henriques, D., Villaverde, A. F., Rocha, M., Saez-Rodriguez, J., & Banga, J. R. (2017). Data-driven reverse engineering of signaling pathways using ensembles of dynamic models. PLOS Computational Biology. https://doi.org/10.1371/journal.pcbi.1005379

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.