Limits...
Diagnostics for stochastic genome-scale modeling via model slicing and debugging.

Tsai KJ, Chang CH - PLoS ONE (2014)

Bottom Line: In doing so, they have traded the ability to have exchangeable, standardized model representation formats, while those that remain true to standardized model representation are faced with challenges in model complexity and analysis.The computer-aided identification revealed specific points of interest such as reversibility of reactions, initialization of species amounts, and parameter estimation that improved a candidate cell's adenosine triphosphate production.We then compared the advantages of our methodology over other modeling techniques such as model checking and model reduction.

View Article: PubMed Central - PubMed

Affiliation: Bioinformatics Program, Taiwan International Graduate Program, Academia Sinica, Taipei, Taiwan; Institute of Biomedical Informatics, National Yang-Ming University, Taipei, Taiwan.

ABSTRACT
Modeling of biological behavior has evolved from simple gene expression plots represented by mathematical equations to genome-scale systems biology networks. However, due to obstacles in complexity and scalability of creating genome-scale models, several biological modelers have turned to programming or scripting languages and away from modeling fundamentals. In doing so, they have traded the ability to have exchangeable, standardized model representation formats, while those that remain true to standardized model representation are faced with challenges in model complexity and analysis. We have developed a model diagnostic methodology inspired by program slicing and debugging and demonstrate the effectiveness of the methodology on a genome-scale metabolic network model published in the BioModels database. The computer-aided identification revealed specific points of interest such as reversibility of reactions, initialization of species amounts, and parameter estimation that improved a candidate cell's adenosine triphosphate production. We then compared the advantages of our methodology over other modeling techniques such as model checking and model reduction. A software application that implements the methodology is available at http://gel.ym.edu.tw/gcs/.

Show MeSH
A list of reactions as seen in the main workspace of the diagnostic application.The highlighted reaction represents the reaction that the Gillespie algorithm has chosen to execute based on reaction weight. The predictive weight column informs the user of the normalized probability of the reaction taking place even if the reactants are not directly available. For example, R_PYK, has a zero reaction weight due to the lack of phosphoenolpyruvate but a predictive weight of 10−5. This value is a result of the availability of glucose and the upstream R_HEX1 reaction.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4219680&req=5

pone-0110380-g006: A list of reactions as seen in the main workspace of the diagnostic application.The highlighted reaction represents the reaction that the Gillespie algorithm has chosen to execute based on reaction weight. The predictive weight column informs the user of the normalized probability of the reaction taking place even if the reactants are not directly available. For example, R_PYK, has a zero reaction weight due to the lack of phosphoenolpyruvate but a predictive weight of 10−5. This value is a result of the availability of glucose and the upstream R_HEX1 reaction.

Mentions: In order to run an initial simulation of the model we needed to populate the model with an initial amount of species and non-zero reaction kinetic law values, two pieces of information not in the original model. We first set all reactions to an arbitrary 0.1 kinetic law value with the intention of refining the species amounts first. With non-zero kinetic law values we were able to calculate reactions weights and use predictive weights to determine species that would have an effect on the eventual production of ATP, even if they were not related to reactions that directly produced ATP. Each time a new metabolite was produced, the predictive weight for all reactions in the model slice were recalculated. A screenshot of the diagnostic application's reaction workspace with reaction and predictive weights is seen in Figure 6.


Diagnostics for stochastic genome-scale modeling via model slicing and debugging.

Tsai KJ, Chang CH - PLoS ONE (2014)

A list of reactions as seen in the main workspace of the diagnostic application.The highlighted reaction represents the reaction that the Gillespie algorithm has chosen to execute based on reaction weight. The predictive weight column informs the user of the normalized probability of the reaction taking place even if the reactants are not directly available. For example, R_PYK, has a zero reaction weight due to the lack of phosphoenolpyruvate but a predictive weight of 10−5. This value is a result of the availability of glucose and the upstream R_HEX1 reaction.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4219680&req=5

pone-0110380-g006: A list of reactions as seen in the main workspace of the diagnostic application.The highlighted reaction represents the reaction that the Gillespie algorithm has chosen to execute based on reaction weight. The predictive weight column informs the user of the normalized probability of the reaction taking place even if the reactants are not directly available. For example, R_PYK, has a zero reaction weight due to the lack of phosphoenolpyruvate but a predictive weight of 10−5. This value is a result of the availability of glucose and the upstream R_HEX1 reaction.
Mentions: In order to run an initial simulation of the model we needed to populate the model with an initial amount of species and non-zero reaction kinetic law values, two pieces of information not in the original model. We first set all reactions to an arbitrary 0.1 kinetic law value with the intention of refining the species amounts first. With non-zero kinetic law values we were able to calculate reactions weights and use predictive weights to determine species that would have an effect on the eventual production of ATP, even if they were not related to reactions that directly produced ATP. Each time a new metabolite was produced, the predictive weight for all reactions in the model slice were recalculated. A screenshot of the diagnostic application's reaction workspace with reaction and predictive weights is seen in Figure 6.

Bottom Line: In doing so, they have traded the ability to have exchangeable, standardized model representation formats, while those that remain true to standardized model representation are faced with challenges in model complexity and analysis.The computer-aided identification revealed specific points of interest such as reversibility of reactions, initialization of species amounts, and parameter estimation that improved a candidate cell's adenosine triphosphate production.We then compared the advantages of our methodology over other modeling techniques such as model checking and model reduction.

View Article: PubMed Central - PubMed

Affiliation: Bioinformatics Program, Taiwan International Graduate Program, Academia Sinica, Taipei, Taiwan; Institute of Biomedical Informatics, National Yang-Ming University, Taipei, Taiwan.

ABSTRACT
Modeling of biological behavior has evolved from simple gene expression plots represented by mathematical equations to genome-scale systems biology networks. However, due to obstacles in complexity and scalability of creating genome-scale models, several biological modelers have turned to programming or scripting languages and away from modeling fundamentals. In doing so, they have traded the ability to have exchangeable, standardized model representation formats, while those that remain true to standardized model representation are faced with challenges in model complexity and analysis. We have developed a model diagnostic methodology inspired by program slicing and debugging and demonstrate the effectiveness of the methodology on a genome-scale metabolic network model published in the BioModels database. The computer-aided identification revealed specific points of interest such as reversibility of reactions, initialization of species amounts, and parameter estimation that improved a candidate cell's adenosine triphosphate production. We then compared the advantages of our methodology over other modeling techniques such as model checking and model reduction. A software application that implements the methodology is available at http://gel.ym.edu.tw/gcs/.

Show MeSH