Most pooling variation in array-based DNA pooling is attributable to array error rather than pool construction error

Eur J Hum Genet. 2007 Apr;15(4):501-4. doi: 10.1038/sj.ejhg.5201768. Epub 2007 Jan 31.

Abstract

Genome-wide association (GWA) approaches are important in complex disease gene mapping studies but are often prohibitively expensive. Array-based DNA pooling has been shown to offer substantial cost savings compared with individual genotyping. This reduced cost potentially brings well-powered GWA studies well within the reach of most laboratories. The main factor, which affects the efficiency of pooling compared with individual genotyping is the magnitude of the pooling error variance. By examining variation between and within pools it is shown that most of the error associated with pooling is attributable to array variation not pooling construction variation (assuming the pools are not small and the pools are accurately constructed). With Affymetrix HindIII 50K arrays used here the array-specific variance is seven times the pooling construction variance. This has important implications for optimal study design for array-based pooling. Given carefully constructed pools, resources should be allocated to increasing the number of arrays per sample rather than to constructing multiple pools.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Analysis of Variance
  • Chromosome Mapping / standards
  • DNA / genetics*
  • Gene Expression Profiling / methods*
  • Gene Expression Profiling / standards
  • Gene Frequency / genetics*
  • Gene Pool*
  • Genetic Variation / genetics
  • Humans
  • Likelihood Functions
  • Linkage Disequilibrium
  • Models, Genetic*
  • Oligonucleotide Array Sequence Analysis
  • Research Design / standards*
  • Sample Size
  • Selection Bias

Substances

  • DNA