Haplotype Mapping Models for Mutant Genes
                 in Nonstationary Population Isolates
                         

                              Ruzong Fan
               Center for Biostatistics in AIDS Research
                     Harvard School of Public Health

      Geneticists are now actively attempting to map disease genes. 
Genetic linkage analysis can locate the region of a gene responsible for a
disease, and haplotype mapping can help to refine that region. Population
isolates such as Finland and Costa Rica provide unique opportunities for
mapping disease genes using reconstructed haplotypes. In the presence of a
nonstationary surrounding population, modeling the dynamics of haplotype
evolution is a challenge to geneticists.

      My research explores haplotype evolution in exponentially growing
population isolates. Simple difference equation models are adequate for
understanding the deterministic balance between selection and mutation
near a disease locus. To comprehend the more subtle stochastic aspects of
this balance requires techniques from the theory of branching processes and
stochastic differential equations. One of the major goals of my work is to
apply these stochastic tools to broaden our intuition about the most
favorable circumstances for haplotype mapping.