Reducing the Genotyping Needed for Case-parents Designs by Pooling DNA Samples

• Activity Number: 72
• Type: Topic Contributed
• Date/Time: Monday, August 9, 2004 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): David M. Umbach*+ and Clarice R. Weinberg
• Companies: National Institute of Environmental Health Sciences and National Institute of Environmental Health Sciences

Abstract

To study disease-related genes, one can genotype affected children and their parents (triads), achieving robustness to genetic population structure. However, for rare alleles, many families carry no copies and are discarded as uninformative. To reduce assays while retaining robustness, we considered design variations in which DNA is pooled before assay. We assume that the assay for each diallelic locus can count the proportion of variant alleles among the 2k present in a pool from k individuals. If DNA from the two parents in each triad is pooled but the child is assayed separately, one can carry out a test that resembles the TDT while reducing assays by one-third. One can also analyze such data via log-linear models. A second design reduces genotyping by two-thirds by randomly matching pairs of triads and assaying one pool of the four parental samples and a second of the two offspring samples. The data can be analyzed with a log-linear modeling approach via the EM algorithm. We study the operating characteristics of these designs under different allele frequencies and modes of inheritance. DNA pooling can be advantageous when specimen volumes are limited or the allele is rare.