__GNTInfer__: A Gene Network reconstruction tool with compound Targets by Integrating multiple microarray datasets and prior information * Version 1.1 * Last updated: January, 2007 !!!Reference * Yong Wang, Trupti Joshi, Xiang-Sun Zhang, Dong Xu, and Luonan Chen. __Recovering gene regulation and identifying compound targets by integrating multiple time course expression datasets and prior information__. ''Manuscript'', 2007. !!!Update * January, 2007: Version 1.1 Update the input file format and improve the algorithm. !!!Method GNTInfer aims to combine computational analysis of multiple microarray datasets and biological experiment results together for inferring gene regulatory network and further identifying compound targets of perturbation experiments. The GNTInfer is developed by extending [GRNInfer] to include other available information derived from expression profile and from published literature so as to recover gene regulations in a more robust and reliable manner, and to incorporate external inputs or perturbations into the formulation so that molecular targets (genes) can be identified in a systematic way. !!!Procedure * __Step 1__: Format the Multiple Microarray Datasets into desired data file (See the attached example file as a reference, notice in the version 1.1 there are some changes in the file format. All the documented regulatory relationship, co-expression relationships, TF DNA binding data, the external perturbation information and multiple time series datasets are formulated in a single file. In them prior information is classifed as two catalogs. The first class contains the interactions with known activating and repression condition (such as some documented edges) and the known interactions without activating or repressing information belong to the second class (such as chIP-chip data). Also perturbation are classified as the known ones (to which gene and at which time point) and unknown ones (need to infer the gene targets). For a detailed example to reconstruct 145-gene regulatory network in yeast is attached with the software. The python script to deal with the file format is available upon request.) * __Step 2__: Open the data file by click the 'open' button and browse the data file location. * __Step 3__: Choose proper parameters. There are three parameters you can choose by altering the default value. Specifically, Lambda: This parameter is used in the inferring algorithm to adjust the sparsity of the structure. The default value is 0.0. Link Strength Threshold: This parameter is used in the control the output file GRN.dot, which can be visualized by the neato tool of software Graphviz. The threshold parameter controls the number of the edge whose strength of link is smaller than Threshold not shown in the network graph. The smaller this parameter, the more edges in the network graph. The default value is set to 1e-6. Co-expression Threshold: This parameter is used to control the role of co-expression information in the inference algorithm. The co-expression is computed based on microarray profile and four methods are combined together to determine the co-expression relationships. * __Step 4__: Computing by click the 'Infer' button when the data file and parameters are ready. * __Step 5__: Checking the results. There are four files generated in the same directory with the software. ** __GRN.dat__: This file records the matrix assessing the strength of edges of the network. i.e. the element located on the ith row and the jth column means the regulatory strength from the jth gene to the ith gene, the sign + means the activation and the sign - means repression. The extra columns record the inferred perturbation strength applied to each gene. ** __GRN.dot__: This file records the information to output the network structure which can be illustrated by software Graphviz. Note that only the edges whose regulatory strength in GRN.dat is larger than Threshold parameter are shown. The software Graphviz can be download from [http://www.graphviz.org] freely. ** __Interaction.txt__: This file list all the inferred links with their activating or repressing conditions and weight. ** __GNTInfer.log__: The log file for the computation. You can check it to make sure there are no error happens. !!!Software This is a beta version of the program for preliminary testing. The program is still under development. * [GNTInfer ver 1.1.zip|GNTInfer/GNTInfer ver 1.1.zip] ---- Category: [Software]