Mapping genes for complex traits using high performance computing
Title: Mapping genes for complex traits using high performance computing
SNIC Project: 343-2002-2315-46
Project Type: SNAC Large
Principal Investigator: Leif Andersson <Leif.Andersson@bmc.uu.se>
Duration: 2002-07-01 – 2002-12-31
Classification: 152400 152306 133202
Keywords: QTL, genetics

Abstract

Many biological traits show a complex genetic inheritance. This means that multiple genes in combination with environmental faxctors control them. But already in 1926, Sax showed that the fruit fly genetic markers can be used to identify genetic factors underlying complex traits (now called QTL or quantitative trait loci. As the genetical map expands rapidly more complete datasets are obtained. This opens up the possibility to explore more complex modes of gene action. The computational demand increases rapidly with an increase in complexity of the analysis model. Likewise there is a rapid increase when considering multiple traits and multiple allelic variants at QTL.Implementation of high performance computing analyses drastically improves the output and efficiency as well as the scientific quality of this kind of research,also our research.The possibility to use high performance computing analyses has allowed us to overcome the obstacle of randomisation testing in multiple QTL models. A large simulation study, which resulted in a procedure now used in our software, was performed at the T3E. This procedure is the first of its kind for significance testing in multiple QTL models including interactions. Data on a large amount of biological traits has/is being collected form two large chicken populations. Analysis of these data will take place in 2002. The analysis of these data is the way to conclude this project and apply developed methodology on real data and we are here applying for time on the Cray T3E to perform these analyses.