Mapping segmental imbalances using comparative genomic hybridization and eigenanalysis

We have tested a new approach to comparative genomic hybridization (CGH) analysis using digital ratio images and eigenanalysis, which allows the recognition of consistent patterns along the chromosomes and discards random (background noise) patterns. We have performed test experiments using genomic DNAs from a patient with a duplication, another with a deletion of a chromosome segment, and a prostate cancer biopsy. Image ratio analysis was performed, and ratio images of the relevant chromosome were subjected to eigenanalysis. The results showed a high-contrast enhancement of the regions corresponding to the unbalanced genomic segment, with clearly defined limits between normal and abnormal fluorescence ratios. The combination of digital ratio images and eigenanalysis allowed the precise mapping of unbalanced regions consistent with other methods of analysis. Because there is no limit to the number of chromosomes that can be analyzed at any one time, the method has the potential of increasing the sensitivity of CGH by reducing the noise component.