Batch high throughput multidimensional scaline (HiT-MDS) for MATLAB/GNU Octave programming language QUICK START: - extract files to directory (which You probably did already), - start MATLAB/GNU-Octave, - change to target directory (cd), - on MATLAB prompt type> embed_genes GNU-Octave prompt type> embed_genes_octave A scatter plot of 1206 points display should appear and form a sandglass shape of temporally up- and down-regulated genes. Final correlation should be around r^2 ~ 0.853. For smaller data sets restart multiple times for getting a best solution. See program code for further information end enjoy the program, Marc Strickert (stricker@ipk-gatersleben.de) FILES distmat.m: helper to calculate Euclidean distance matrix stdscatter.m: helper to create a standardized orientation of the scatter plot by PCA embed_genes_octave.m: the caller for the GNU-Octave version hitmds_cpu2_sparse_octave.m: the GNU-Octave implementation of HiT-MDS embed_genes.m: the caller for the MATLAB version hitmds_cpu2_sparse.m: the MATLAB implementation of HiT-MDS genes_endo_4824.dat: 14-dim data of regulated genes of developing barley endosperm tissue see: http://www.biomedcentral.com/1471-2105/8/165 gp_license.txt: Licence text (GPLv2) Further link: http://hitmds.webhop.net/ Leibniz-Institute of Plant Genetics and Crop Plant Research Gatersleben http://www.ipk-gatersleben.de/ This work is kindly founded by the Ministry of Cultural Affairs of Sachsen Anhalt Grant: XP3624HP/0606T Sun May 17 11:07:01 2009 // EOF