Objective Analysis of Chromosome Banding Patterns

This project aims to develop and commercialize significantly improved software for chromosome bandianalysis and anomaly detection. Used with modern computerized chromosome analysis, this will allow ocharacterization and convenient display of chromosome banding patterns, and quantitative comparisonnormal and abnormal prototypes and to each other. This will help to detect and quantify sub-visibleparticularly with high resolution banding. This will significantly increase the ability of automatedinstruments to evaluate chromosome alterations in normal and neoplastic mammalian cells. Specificallfrequently domain eigenanalysis, a new image analysis technique developed at the Baylor College of Mdecompose chromosome images into components, selecting those that reflect relevant characteristics owhile eliminating components reflecting preparatory artifact, inter- chromosome variation, and otherWe will extend the eigenanalysis software so that it can (1) interactively test chromosomes for specautomatically screen chromosomes for a specified group of abnormalities, (3) automatically classifycomparing each to the prototype of each homolog type, (4) compare homologs to each other to detect sand (5) develop prototypical chromosome images for groups of people manifesting genetically linked dthese with normals. We will implement this on PSII's Genetiscan commercial automatic karyotyping wor