文档介绍：Empirical Data Indicates a Primarily Additive ic Model for Expressional QTLs
Ke Hao Andrew Kasarskis
Research ics, Rosetta Inpharmatics Research ics, Rosetta Inpharmatics
Merck Research Lab, Merck & Co. Inc. Merck Research Lab, Merck & Co. Inc.
401 Terry Avenue N, Seattle, WA 98109, United States 401 Terry Avenue N, Seattle, WA 98109, United States
Ke_Hao@ Andrew_Kasarskis@
Abstract— Additive and dominant/recessive traits are more ic models would worsen the multiple testing
frequently encountered in laboratory and natural populations, penalty. Here we attempt to empirically evaluate the
but the fraction of loci segregating dominant alleles has not frequency of ic models. Clinical phenotypes may not
been investigated with experimental data on many traits in the serve our aim well because we only have a handful of such
human population. Existing data and recent theoretical work traits underwent GWAS, and most of the raw data is not public
suggest that the majority of ic variation is additive, but the available. Instead, we taken advantage of gene-mapping on
distribution of the fit of individual loci across the genome has expression quantitative trait loci (eQTL) 7,8, where tens
not been surveyed. Here we show a smooth distribution thousand of GWAS were conducted in parallel. SNPs
ranging from purely additive to over-dominant/recessive associ