Rapid advances in biological technology have radically changed the amount and type of information available for the location of genes. Statistical methodology needs to keep pace with the advances in molecular technology so that the information can be exploited to the fullest possible extent. In this project, we propose to greatly expand existing statistical methodology for the Iocation of binary traits. Our proposed methodology will be able to model evolutionary situations including: multiple genes, varying penetrance, and simple environmental effects. Not only will we provide methods for analyzing these data, but we propose to provide a free computer program that will make these methods accessible to the larger scientific community. We are proposing to develop methodology for traits that are binary in nature, that is these traits are either present or absent. For example, resistance to disease or resistance to pests is often scored as present or absent. Understanding the underlying genetic factors influencing plant and animal natural resistance to disease and insects is critical in the understanding of evolutionary forces. Natural selection acts through survival and fitness, and the fitness of individuals can be affected by resistance factors. These binary traits, while simple in their description, are not necessarily simple in their expression. The mechanisms underlying resistance may be complex, involving multiple genes and varying levels of penetrance. Resistance may occur in certain combinations of alleles at multiple loci. For example, it has been demonstrated in Maize that there are many genetic factors affecting resistance to Fusarium stalk rot (McMullen and Simcox 1995). In addition to complex genetic factors affecting resistance, environmental factors may also play a significant part in the expression of resistance. For examp le, a disease may attack a point in a signal transduction cascade (Staskawicz, Ausubel, Baker, Ellis, and Jones 1995). There are several gene products involved in the cascade and a change in the gene product at any point in the cascade may affect the outcome (disease present or absent). In this model, an environmental trigger, the presence of a specific pathogen, is assumed (Staskawicz, Ausubel, Baker, Ellis, and Jones 1995). Without the ability to directly model these complex factors, the understanding of the influence of these factors in the observed phenotype will be limited. Thus, it is clear that statistical methods must exist that are capable of discerning the location and effect of multiple genes for a single binary trait in an environmental setting. In these complex situations, particular care must be taken in fitting the models. We will be using maximum likelihood approaches to develop and fit the statistical models. The specific objectives of this grant are: 1.To develop a solid methodology for the analysis of binary traits resulting from single gene and multigenic models with varying levels of penetrance. 2.To develop solid methodology for the analysis of binary traits incorporating environmental factors in single gene and multigenic models with varying levels of penetrance. 3.To provide an easy to use, free computer program to perform the analyses developed by objectives 1 and 2. References McMullen, M. and K. Simcox (1995). Genomic organization of disease and insect resistance genes in maize. Molecular Plant-Microbe Interactions 8, 811-815. Staskawicz, B., F. Ausubel, B. Baker, J. Ellis, and J. Jones (1995). Molecular genetics of plant disease resistance. Science 268, 661-667.
