Large-scale executable biology using rapid integration of computational models: ee

A1 Originalartikel i en vetenskaplig tidskrift (referentgranskad)


Interna författare/redaktörer


Publikationens författare: Vladimir Rogojin, Ion Petre
Förläggare: Academy of Sciences of Moldova
Publiceringsår: 2016
Tidskrift: Computer Science Journal of Moldova
Tidskriftsakronym: CSJM
Volym: 24
Nummer: 1 (70)
Artikelns första sida, sidnummer: 118
Artikelns sista sida, sidnummer: 135


Abstrakt

We plan to develop a systematic framework for assembling large-
scale computational biological models by reusing and combining al- ready existing modelling efforts. Our goal is to build a software plat- form that will compile large-scale biomodels through successive inte- grations of smaller modules. The modules can be arbitrary executable programs accompanied by a set of (I/O) interface variables; they may also have an internal structure (such as a metabolic network, interac- tion network, etc.) that yields its executable part in a well defined way. Firstly, wherever possible, modules with the compatible inter- nal structure will be joined by combining their structure and by pro- ducing new larger executablemodules (like, combining two metabolic networks, etc.). Then, irrespective of the underlying internal structure and modelling formalisms, all the modules will be integrated through connecting their overlapping interface variables. The resulting com- posed model will be regarded as an executable program itself and it will be simulated by running its submodules in parallel and synchroniz- ing them via their I/O variables. This composed model in its turn can also act as a sub-module for some other even large composite model. Themajor goal of this project is to deliver a powerful large-scalemod- eling methodology for the primary use in the fields of Computational Systems Biology and Bioinformatics.
We plan to develop a systematic framework for assembling large-scale computational biological models by reusing and combining already existing modelling efforts. Our goal is to build a software platform that will compile large-scale biomodels through successive integrations of smaller modules. The modules can be arbitrary executable programs accompanied by a set of (I/O) interface variables; they may also have an internal structure (such as a metabolic network, interaction network, etc.) that yields its executable part in a well defined way. Firstly, wherever possible, modules with the compatible internal structure will be joined by combining their structure and by producing new larger executable modules (like, combining two metabolic networks, etc.). Then, irrespective of the underlying internal structure and modelling formalisms, all the modules will be integrated through connecting their overlapping interface variables. The resulting composed model will be regarded as an executable program itself and it will be simulated by running its submodules in parallel and synchronizing them via their I/O variables. This composed model in its turn can also act as a sub-module for some other even large composite model. The major goal of this project is to deliver a powerful large-scale modeling methodology for the primary use in the fields of Computational Systems Biology and Bioinformatics.


Nyckelord

Computational Biomodeling, Executable Biology, fit-preserving refinement, Model integration, Multi-scale modeling, quantitative model refinement, Whole-cell modeling


Dokument


Senast uppdaterad 2019-21-11 vid 05:09