Support is requested for five years to conduct theoretical and computer simulation studies which arise out of attempts to apply current models and methods of quantitative genetic analysis to human data. The issues we hope to address are: 1. How far we can detect the genetic control of sensitivity to indices of the environment (""""""""genotype x environment interaction""""""""); 2. The simultaneous genetic analysis of multiple variables including a study of: errors arising as a result of misspecifying models for environmental indices; issues in the design and analysis of studies designed to measure genetic covariation between measures; 3. statistical difficulties associated with trying to apply path models to discontinuous data; 4. errors which can occur in trying to detect genes of large effect against a background of polygenic variation. Each of these issues will be addressed by simulating data under a different mechanism of inheritance. By fitting the """"""""right"""""""" model and attempting to devise simple statistical tests of particular hypotheses we shall determine the reliability with which particular effects may be detected and efficient estimates of the population parameters be recovered. By fitting """"""""wrong"""""""" models, we shall find out how often investigators could come to the wrong conclusions and base further developments on false explanation of their data.
Showing the most recent 10 out of 49 publications
