1 Epidemiology, Biostatistics and Biodemography, Department of Public Health, Det Sundhedsvidenskabelige Fakultet, SDU2 Human Genetics, Department of Clinical Research, Det Sundhedsvidenskabelige Fakultet, SDU3 MRC Epidemiology Unit, The Strangeways Research Laboratory, Worts Causeway,4 Epidemiology, Biostatistics and Biodemography, Department of Public Health, Det Sundhedsvidenskabelige Fakultet, SDU
In human genetics, large-scale data are now available with advances in genotyping technologies and international collaborative projects. Our ongoing study of obesity involves Affymetrix 500k genechips on approximately 7000 individuals from the European Prospective Investigation of Cancer (EPIC) Norfolk study. Although the scale of our data is well beyond the ability of many software systems, we have successfully performed the analysis using the statistical analysis system (SAS) software. Our implementation trades memory with computing time and requires moderate hardware configuration. By using such an established system, it extends some earlier discussions in a more constructive and accessible way. We report our findings and give some recommendations with SAS. We also compare briefly with alternative implementations. Our work is relevant to researchers conducting analysis of large-scale data in general, and genomewide association studies in particular.
Main Research Area:
Third International Conference on Intelligent Computing, 2007