<body bgcolor="#FFFFFF"> <h1>MetaGeneAlyse</h1> <h2>Analysis of gene expression/metabolite data</h2> <h3>Carsten Daub, Sebastian Kloska, and Joachim Selbig</h3> <br> <a href="http://www.mpg.de/english/" target=_blank>Max Planck Society</a>&nbsp;&nbsp;&nbsp; <a href="http://www.mpimp-golm.mpg.de/" target=_blank>MPI-MP</a>&nbsp;&nbsp;&nbsp; <a href="http://www.mpimp-golm.mpg.de/bioinformatics" target=_blank>Bioinformatics - Central Infrastructure Group</a>&nbsp;&nbsp;&nbsp; <a href="http://www.mpimp-golm.mpg.de/2278/impress" target=_blank>Imprint</a>&nbsp;&nbsp;&nbsp; <a href="http://www.mpimp-golm.mpg.de/2042773/Privacy-Policy" target="_blank">Privacy policy</a> <br><br><br> <div align="justify"> <p> Within this system you are able to upload a dataset containing gene expression and/or metabolite data. </p> <p> Each uploaded dataset is assigned an 8-digit ID. This file ID is used to access your dataset for analysis. </p> <p> The <i>MetaGeneAlyse</i> server calculates the distance between each gene in a dataset with every other gene in pairwise comparisons. For N genes, a so-called distance matrix of size N x N is calculated based on one of a set of similarity measures. The matrix is symmetric and the diagonal elements are zero. This matrix also gets an ID and can be downloaded or one of a set of cluster algorithms can be applied to the matrix. </p> <p> The result of the clustering comes in the form of a dendrogram and can be displayed in different formats. Optionally the dendrogram can be cut at a certain hight. The associations of genes to clusters are displayed in a table. </p> <p> The k-means clustering algorithm can also be applied to the uploaded dataset. </p> </div> </div> </body>