GeRNet: A Framework for Inference, Visualization and Manipulation of Gene Regulatory Networks based on Association Rules

GALLO, CRISTIAN ANDRÉS; CARBALLIDO, JESSICA ANDREA; PONZONI, IGNACIO

Rosario

Congreso; 4to. Congreso Argentino de Bioinformática y Biología Computacional (4CAB2C) y 4ta. Conferencia Internacional de la Sociedad Iberoamericana de Bioinformática (SolBio); 2013

A2B2C

Background Gene regulatory networks (GRNs) play an important role in the progression of life phenomena such as cell cycling, developmental biology, aging, and the progressive and recurrent pathogenesis of complex diseases, among others. The amount of genome-wide gene expression time series data is becoming increasingly available, providing the opportunity to reverse engineer the time-delayed gene regulatory networks that govern the majority of these molecular processes. In this context, data mining methods constitute suitable approaches for performing the inference of the relational structures of a GRN. In particular, these reconstruction strategies can be beneficed from the application of association rules extraction techniques [1]. Method The aim of the research presented here consists on the reconstruction of gene regulatory networks based on association rules from multiple genome-wide microarray time series datasets given as input. In this regard, a new user-friendly software tool called GeRNet was developed. The software integrates a combinatorial optimization procedure of gene profile classifiers [2] for the inference of potential time-delay association rules between genes. Association rules establish causal links between two genes, where the semantics and the interpretation of the relation depend of the input data and on the rule type inferred. The platform also offers a biclustering algorithm [3] for the inference of additional relationships that may not be captured by the main inference algorithm, thus enhancing the overall inference capabilities. Additionally, the framework provides with several features that are summarized as follows: Data handling. All data, being the set of expression data matrices loaded from external files or the gene regulatory network, are organized in a tab structured that is depicted in the right panel of the graphical user interface (Figure 1). The left panel shows the list of gene regulatory associations presented in the network. All the work performed with the software can be saved as a project form and restored later. Data pre-processing. The input data can be any set of CSV files including annotations of genes and conditions. The software provides with a tool for inferring the missing values of the data. Visualization and Manipulation. The expression matrices are displayed in the Datasets tab and can be viewed as heatmaps or as numerical matrices. Annotations of the conditions run along the top whereas the annotations of the genes are listed on the left hand side. On the other hand, the gene regulatory network is displayed in the GRN View tab as a graph, where each gene is represented as a green circle vertex and the rules that related them are directed edges drawn accordingly to the type of the rule. The visualization mechanism allows for several different views of the network, which can be selected regarding the time-delay, rule type, or gene names of interest. Additionally, the annotations along the network can be selectively adapted to provide the desire information. It also provides with several types of graph layouts that automatically rearrange the genes in the network. On the left side of the GRN View there is a tool bar aimed to interact with the network. It contains tools that allow zooming, panning, rotating, or translating the GRN. Moreover, it allows the edition of the network by means of inserting, deleting and selecting genes and interactions. All the results and views can be exported in several formats for further usages on presentations, papers, etc. Conclusions In this work, we have introduced a software platform called GeRNet. The software represents a user-friendly framework that integrates two recently published algorithms, providing the opportunity to reverse engineer time-delayed gene regulatory networks based on association rules. It also offers several features for data handling, visualization and manipulation of gene regulatory networks, creating a complete bioinformatics environment for the research biologists. However, desired features like integration with external biological knowledge such as KEGG and Gene Ontology are put in forward for further releases. Availability The GeRNet software is freely available at http://lidecc.cs.uns.edu.ar and runs on all operating systems with a Java Virtual Machine.