International Mammalian Genome Society

The 16th International Mouse Genome Conference (2002)

Plenary Presentations * Oral Presentations *
Poster Presentations: Complex Genetics and Disease Modifiers * Developmental Genetics * Functional Genomics * Gene Discovery * Genetic Manipulations to Alter Gene Function * Mouse Models: Human Disease and Pharmacogenetics * Sequence Annotation and Comparative Analysis of Genomes *
Attendees * Sponsors * Table of Contents * Photographs * Awards

POSTER 18 - QTL MAPPING WITH MICROARRAY EXPRESSION DATA

KF Manly
Roswell Park Cancer Institute

1) Wang J, 2) Shou S, 2) Qu Y, 2) Chesler E, 2) Lu L, 3) Hsu HC, 3) Mountz JD, 2) Threadgill D, 4) Williams RW
1) Roswell Park Cancer Institute, 2) Univ Tennessee, Memphis, 3) University of Alabama, Birmingham, 4) University of North Carolina

Using high density oligonucleotide arrays (Affy U74Av2) we measured expression of 12422 transcripts in forebrains of 24 lines of adult BXD mice (including parents and F1).Experimental noise was reduced by pooling and replication: 3 cases/array and 1 to 4  arrays/line. Log-transformed expression data (average difference between PM and MM probes) for each array were normalized to standard mean and SD before replicate array data were averaged. Variation in transcript level was mapped by least-squares regression against genotypes of 508 non-redundant BXD markers, using custom software written in C and Python. Empirical p-values for each transcript/locus association were calculated by permutation tests with up to 1000000 permutations.Only the most significant marker association was retained for each transcript. These were ordered by p-value and tested according to Benjamini and Hochberg at false discovery rate = 0.2. This test identified 74 QTLs, distributed across almost all chromosomes.Repeating this procedure with data from individual probes (about 400,000) declared 576 probe-locus combinations, representing 339 transcripts, as significant. 79% of these were derived from PM probes. Individual probes from the same probeset differed greatly in their association with any marker, and for most transcripts only one probe (usually a PM probe) achieved significance when tested individually.*Benjamini and Hochberg (1995) Controlling the False Discovery Rate: a practical and powerful approach to multiple testing. J Royal Statistical Society B 57(1): 289-300Supported by: Dunavant Chair, Univ Tennessee Health Science Center, and the Human Brain Project (MH62009, NIMH)