Introduction
Genomics and proteomics data are typically analyzed by hierarchical clustering, followed by visualization with heatmaps [1-3]. Various algorithms have been implemented in the data clustering procedure [4]. The visualization of clustered data includes tree-based hierarchical clustering patterns and heatmaps of experimental values [5]. Simultaneously carrying out clustering and visualization in a single platform provides a convenient tool for choosing an appropriate clustering algorithm and finding patterns in the resulting heatmaps. Previously, bioinformaticists used programmable tools, such as R and Matlab, and commercial data-mining packages to analyze their data. A simple and integrated program will allow experimental scientists to intuitively identify meaningful patterns from a large dataset without requiring knowledge of scripting computer languages or statistical theory.
Herein, we introduce a user-friendly tool, QCanvas, which integrates diverse clustering algorithms and an interactive heatmap display interface (Fig. 1). This program directly imports raw experimental data in a matrix format and displays these data in a heatmap. Various clustering methods can be applied to two-dimensional data, with the real-time generation of clustered heatmaps. Furthermore, subsets of heatmap data can be selectively displayed, based on user-defined filters. QCanvas is an easy-to-use and powerful tool for fast data analysis and interpretation by bench scientists. Without any knowledge of scripting languages and without any graphics-editing software, one can generate and customize tree-clustered heatmaps with high-quality graphics.