This project will implement a novel approach to understanding the relationship between genetic variation and its effects on the phenotype that is to the genotype-phenotype (GP) map. Genotypes exist and are inherited in a discrete space convenient for many sorts of analyses, but the causation of the important phenomena such as disease status and natural selection takes place in a continuous phenotype space whose relationship to the genotype space is only dimly grasped. Direct study of genomes alone, with minimal reference to phenotypes is insufficient to understand why some individuals are sick and some are healthy. This project takes an integrative approach to the study of the GP map, combining both genetic and phenotypic studies that potentially reinforce each other. Wing shape in Drosophila melanogaster is an excellent model system for this because the phenotype can readily be characterized in many dimensions, there is substantial natural genetic variation, and because of the genetic control possible in Drosophila.
Aim 1 of the project is to characterize the effects of variation in gene expression at a large number and variety of genes by controlled manipulations of gene expression.
Aim 2 will develop a model of wing development that incorporates these results and produces a complete wing phenotype, something that has only rarely been attempted.
Aim 3 is to identify naturally occurring genetic variation in wing phenotype using association mapping of multivariate effects on wings. By simultaneously considering all phenotypes in the analysis, the problems of mapping multiple traits one at a time will be avoided.
In Aim 4 emerging generalizations about the causes of variation in wing form from Aims 1-3 will be tested by predicting how a population should respond to artificial selection, then testing these predictions in a series of experiments. If a coherent picture of how genetic variation produces phenotypic variation emerges from these experiments, the combination of genetic manipulations, detailed modeling, and characterization of natural variation and its effects can readily be applied to mammalian systems and ultimately to humans. If no such picture emerges, the precise ways in which predictions fail will focus attention on those aspects of the GP map that we do not yet understand.
This work will test whether detailed phenotypic data can be combined with existing genomic data to provide improved predictions of important events, such as disease status or outcome. Current research emphasizes using genotypic data alone for prediction. Recent results show that the nature of genetic causation is very complex, and highly influenced by the environment, strongly limiting the efficacy of a genes-alone approach to health. These problems may be lessened if genomic data is combined with a comprehensive characterization of the phenotype and detailed knowledge of how genetic variation affects those phenotypes.
Showing the most recent 10 out of 11 publications
