A system for integrative multi-dimensional analysis of genomes

¹Raj Chari, ¹Bradley P Coe, ²Calum MacAulay, ¹Wan L Lam
¹British Columbia Cancer Research Centre, Dept. of Cancer Genetics and Developmental Biology, 675 West 10th Avenue, Vancouver, BC, V5Z 1L3, Canada, ²British Columbia Cancer Research Centre, Dept. of Cancer Imaging, 675 West 10th Avenue, Vancouver, BC, V5Z 1L3, Canada

Background: Advances in array based technologies have enabled high throughput genome wide measurement of gene dosage, genetic polymorphism, epigenetic status, and gene expression pattern. Moreover, datasets are emerging for which samples are profiled for multiple features; for example, combinations of gene dosage, loss of heterozygosity, DNA methylation and gene transcript levels. However, there are currently no software tools that facilitate the combination of these complementary data sets to be analyzed in a unified environment.

Objective: To develop software to organize,visualize and analyze multi-dimensional datasets.

Methods: We have established a software package in Java which uses a MySQL database for storage of data and results, and employs the statistical package R for analysis. This new software platform is called SIGMA² for System for Integrated Genomic Microarray Analysis Version 2. The program is developed in Java to facilitate use across all operating systems.

Results: A secure, searchable database has been established to facilitate the storage and optional sharing of genomic data. SIGMA² is highly versatile, having the ability to view data from a variety of commercial and custom microarray platforms. For array based gene dosage analysis (comparative genomic hybridization), multiple visualizations, including signal ratio value and frequency plots, are available at different magnifications, algorithms for automated data segmentation/analysis, and linkage to other datasets. For example, we have incorporated displays for loss of heterozygosity and gene expression data, as well as analysis tools for correlation of gene dosage and gene expression data. SIGMA² is also designed for ease of extension so that new data types can be handled effortlessly and additional algorithms are simple to incorporate.

Conclusion: We have developed a system to perform integrative genomic analysis. Such tools will be necessary for the analysis and interpretation of high-throughput multi-dimensional datasets.