Объектно-ориентированная архитектура информационной системы реконструкции генных регуляторных сетей

Abstract

Описана архитектура информационной системы, предназначенной для решения задач реконструкции генных регуляторных сетей, в основу которой положен объектно-ориентированный подход, определяющий ее открытость, универсальность и расширяемость. Предложен сценарий реконструкции, в котором осуществляется оптимизация пространства поиска значений параметров S-системы.

Описано архітектуру інформаційної системи, призначеної для вирішення завдань реконструкції генних регуляторних мереж, в основу якої закладено об'єктно-орієнтований підхід, який визначає її відкритість, універсальність і розширюваність. Запропоновано сценарій реконструкції, в якому проводиться оптимізація простору пошуку значень параметрів S-системи.

Introduction. Insufficient level of understanding of the nature of regulation and functional mechanisms of gene regulatory networks does not allow to build their mathematical models, based on the fundamental laws of component’s interaction. Now, many different models and methods of gene regulatory reconstruction are developed, which have the advantages and disadvantages. At the choice of descriptive model it is necessary to consider the fact, that mathematical models, as a rule, have their own structure and a number of parameters, which need to be identified. A large number of computational methods are developed, for structuralparametric model’s identification. The majority of them have increased resistance to noise and uncertainty contained in the initial data. The presence of this property is real for the selection of a computational method, used for solving the reconstructing problem of the gene regulatory network based on the gene expression data. Purpose. The purpose of this work is the development of the information system architecture for the gene regulatory network reconstruction, based on the object-oriented approach. Method. The authors used the method of object-oriented design for the developing of this information system. Results. The architecture of the information system for the gene regulatory networks reconstruction, based on the objectoriented approach is proposed. The S-system is applied as a computational model. The parameters and structure are calculated using the clonal selection algorithm. The gene expression profiles are used as an input data. The developed system includes four basic components: the data source, the model, the solution converter and the identification method. The scenario of solving gene network reconstruction problem is developed. In addition, an iterative algorithm for the space optimization search of the computational model parameter values is implemented in this scenario. Conclusion. The developed architecture is open, so that allows to add or replace the separate components by expansions. Further researches suggest to expand the range of used models, such as radial-base network model and wavelet-neural network model, as well as the system for the gene expression programming. In the future, we are planning to implement the addition of new evolutionary algorithms to the information system. In such a way, the work of the evolutionary operators by the development of new scenarios for solving the gene regulatory networks reconstruction problems can be improved.