Research On Microarray Data Analysis And Application Of Microarray On Neurobiology

1. A novel integration and annotation method for microarray dataset. The distributed nature of biological database poses a major challenge to interpretation of large genome-scale dataset emerged from microarray studies. Though there have been a few efforts to address this problem based on warehousing integrating technology, the limited annotation types and out-of-date information contained in these methods largely restrict researchers to further explore the biological knowledge in large microarray datasets. In this paper, we presented an attractive alternative to the existing solution to microarray data annotation combining warehousing technology with link integration and developed a standalone microarray annotation application, Arrayannote. Our results show that Arrayannote can provide up-to-date and systematical annotation information and have unique feature to facilitate post-clustering analysis.2. Visualization of large-scale microarray dataset by an SOM based method. This paper assessed an efficient visualization methodology based on self-organization map (SOM) for large scale microarray data analysis. This approach utilizes component planes, a two-dimensional display consisting of mosaic component, to create specific representations for each sample data, then organizes these component planes into one single graphic sheet according to sample-to-sample relationships, thus proving an readily exploration of the underlying data features by displaying both the natural classes in samples and correlated patterns in the genes simultaneously. Beneficial potentials of this approach were highlighted by processing publicly available large-scale expression data originating from normal human organs as well as human leukaemia study.