Limits...
Exhaustively characterizing feasible logic models of a signaling network using Answer Set Programming.

Guziolowski C, Videla S, Eduati F, Thiele S, Cokelaer T, Siegel A, Saez-Rodriguez J - Bioinformatics (2013)

Bottom Line: We show that, if experimental error is taken into account, there are thousands (11 700) of models compatible with the data.Despite the large number, we can extract structural features from the models, such as links that are always (or never) present or modules that appear in a mutual exclusive fashion.We find 91 behaviors across the 11 700 models and we suggest new experiments to discriminate among them.

View Article: PubMed Central - PubMed

Affiliation: École Centrale de Nantes, IRCCyN UMR CNRS 6597, 44321, Nantes, France.

ABSTRACT

Motivation: Logic modeling is a useful tool to study signal transduction across multiple pathways. Logic models can be generated by training a network containing the prior knowledge to phospho-proteomics data. The training can be performed using stochastic optimization procedures, but these are unable to guarantee a global optima or to report the complete family of feasible models. This, however, is essential to provide precise insight in the mechanisms underlaying signal transduction and generate reliable predictions.

Results: We propose the use of Answer Set Programming to explore exhaustively the space of feasible logic models. Toward this end, we have developed caspo, an open-source Python package that provides a powerful platform to learn and characterize logic models by leveraging the rich modeling language and solving technologies of Answer Set Programming. We illustrate the usefulness of caspo by revisiting a model of pro-growth and inflammatory pathways in liver cells. We show that, if experimental error is taken into account, there are thousands (11 700) of models compatible with the data. Despite the large number, we can extract structural features from the models, such as links that are always (or never) present or modules that appear in a mutual exclusive fashion. To further characterize this family of models, we investigate the input-output behavior of the models. We find 91 behaviors across the 11 700 models and we suggest new experiments to discriminate among them. Our results underscore the importance of characterizing in a global and exhaustive manner the family of feasible models, with important implications for experimental design.

Availability: caspo is freely available for download (license GPLv3) and as a web service at http://caspo.genouest.org/.

Supplementary information: Supplementary materials are available at Bioinformatics online.

Contact: santiago.videla@irisa.fr.

Show MeSH
Experiments to discriminate more relevant GTTs. Both experiments generate the same output in each GTT. Stimuli not shown are inactive, inhibitors not shown are absent and readouts not shown have the same value inthree GTTs
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3753570&req=5

btt393-F6: Experiments to discriminate more relevant GTTs. Both experiments generate the same output in each GTT. Stimuli not shown are inactive, inhibitors not shown are absent and readouts not shown have the same value inthree GTTs

Mentions: Finally, we have investigated the space of experiments to identify the simplest ones (i.e. minimal number of stimulations and inhibitions), which maximize the pairwise differences between the optimal and the two most common GTTs. These three GTTs differ pairwise in either one or two readouts among , and only 192 experiments generate two differences. Out of these 192 experiments, we identified eight experiments with minimal number of stimulations, and among them, we selected the ones with minimal number of inhibitions (Fig. 6). We noted that the two experiments found generate the same output over the readouts. Thus, in contrast to the seven experiments needed to discriminate among all GTTs, only one experiment is required to discriminate between the optimal and the two most common GTTs.Fig. 6.


Exhaustively characterizing feasible logic models of a signaling network using Answer Set Programming.

Guziolowski C, Videla S, Eduati F, Thiele S, Cokelaer T, Siegel A, Saez-Rodriguez J - Bioinformatics (2013)

Experiments to discriminate more relevant GTTs. Both experiments generate the same output in each GTT. Stimuli not shown are inactive, inhibitors not shown are absent and readouts not shown have the same value inthree GTTs
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3753570&req=5

btt393-F6: Experiments to discriminate more relevant GTTs. Both experiments generate the same output in each GTT. Stimuli not shown are inactive, inhibitors not shown are absent and readouts not shown have the same value inthree GTTs
Mentions: Finally, we have investigated the space of experiments to identify the simplest ones (i.e. minimal number of stimulations and inhibitions), which maximize the pairwise differences between the optimal and the two most common GTTs. These three GTTs differ pairwise in either one or two readouts among , and only 192 experiments generate two differences. Out of these 192 experiments, we identified eight experiments with minimal number of stimulations, and among them, we selected the ones with minimal number of inhibitions (Fig. 6). We noted that the two experiments found generate the same output over the readouts. Thus, in contrast to the seven experiments needed to discriminate among all GTTs, only one experiment is required to discriminate between the optimal and the two most common GTTs.Fig. 6.

Bottom Line: We show that, if experimental error is taken into account, there are thousands (11 700) of models compatible with the data.Despite the large number, we can extract structural features from the models, such as links that are always (or never) present or modules that appear in a mutual exclusive fashion.We find 91 behaviors across the 11 700 models and we suggest new experiments to discriminate among them.

View Article: PubMed Central - PubMed

Affiliation: École Centrale de Nantes, IRCCyN UMR CNRS 6597, 44321, Nantes, France.

ABSTRACT

Motivation: Logic modeling is a useful tool to study signal transduction across multiple pathways. Logic models can be generated by training a network containing the prior knowledge to phospho-proteomics data. The training can be performed using stochastic optimization procedures, but these are unable to guarantee a global optima or to report the complete family of feasible models. This, however, is essential to provide precise insight in the mechanisms underlaying signal transduction and generate reliable predictions.

Results: We propose the use of Answer Set Programming to explore exhaustively the space of feasible logic models. Toward this end, we have developed caspo, an open-source Python package that provides a powerful platform to learn and characterize logic models by leveraging the rich modeling language and solving technologies of Answer Set Programming. We illustrate the usefulness of caspo by revisiting a model of pro-growth and inflammatory pathways in liver cells. We show that, if experimental error is taken into account, there are thousands (11 700) of models compatible with the data. Despite the large number, we can extract structural features from the models, such as links that are always (or never) present or modules that appear in a mutual exclusive fashion. To further characterize this family of models, we investigate the input-output behavior of the models. We find 91 behaviors across the 11 700 models and we suggest new experiments to discriminate among them. Our results underscore the importance of characterizing in a global and exhaustive manner the family of feasible models, with important implications for experimental design.

Availability: caspo is freely available for download (license GPLv3) and as a web service at http://caspo.genouest.org/.

Supplementary information: Supplementary materials are available at Bioinformatics online.

Contact: santiago.videla@irisa.fr.

Show MeSH