Whole-genome sequences and multiple trait phenotypes from large numbers of individuals will soon be available in many populations. Well established statistical modeling approaches enable the genetic analyses of complex trait phenotypes while accounting for a variety of additive and non-additive genetic mechanisms. These modeling approaches have proven to be highly useful to determine population genetic parameters as well as prediction of genetic risk or value. We present a series of statistical modelling approaches that use prior biological information for evaluating the collective action of sets of genetic variants. We have applied these approaches to whole genome sequences and a complex trait phenotype resistance to starvation collected on inbred lines from the Drosophila Genome Reference Panel population. We identified a number of genomic features classification schemes (e.g. prior QTL regions and gene ontologies) that provide better model fit and increase predictive ability of the statistical model for this trait.
Original language
English
Publication year
18 Aug 2014
Number of pages
5
Publication status
Published - 18 Aug 2014
Event
10th World Congress on Genetics Applied to Livestock Production (WCGALP) - The Westin Bayshore, 1601 Bayshore Drive, Vancouver, BC V6G 2V4, Vancouver, Canada Duration: 17 Aug 2014 → 22 Aug 2014 Conference number: 10th
Conference
Conference
10th World Congress on Genetics Applied to Livestock Production (WCGALP)
Number
10th
Location
The Westin Bayshore, 1601 Bayshore Drive, Vancouver, BC V6G 2V4
Country
Canada
City
Vancouver
Period
17/08/2014 → 22/08/2014
Research areas
Genomic Feature Models, Whole-genome sequences, Data integration