Stanford Biomedical Informatics

Classification of Expression Arrays Version 1.0

Analysis Tools Classification (Discriminant Analysis Implementation) Clustering (K-means) Visualization (Principal Components Analysis) Home Supplement/Sample Data Documentation References Who we are

K-means Clustering Cluster: genes or arrays File: Distance Metric: Euclidean Manhattan Correlation Based Scale for Variance Scale for Covariance(Mahalonabis) Number of Clusters: Number of Iterations: How do you want to deal with missing values? Zero Fill Average Over Genes Average Over Arrays Indicate, how, and in which order you would like to transform the data, if at all Step 1: Do Nothing Log Data Row Normalize Column Normalize Row Rank Column Rank Step 2: Do Nothing Log Data Row Normalize Column Normalize Row Rank Column Rank Step 3: Do Nothing Log Data Row Normalize Column Normalize Row Rank Column Rank