The project's general aim is to bring the tools and techniques of modern statistics to bear on a variety of problems from genetics and molecular biology. The specific objectives of this project are (i) to develop a suitable statistical model and method of analysis for determining which of a given set of DNA markers come from the same chromosome, based on data obtained from scoring hybrid cell lines; (ii) to extend and refine statistical methods designed to ascertain overlap of cloned DNA fragments; (iii) to develop accurate statistical methods for recognizing the resemblance of a new protein to a known structure or structural motif, on the basis of its amino acid sequence; (iv) to extend the theory of linear invariants for use in general phylogenetic inference; (v) to derive statistical models for the evolution of viral genomes; and (vi) to develop a statistical model for the evolution of certain repetitive DNA elements in human populations. The aim of this project is to develop and use statistical methods for analysing a variety of forms of molecular genetic data. Included in the scope will be data collected on individuals, on single chromosomes, on small cloned DNA fragments, on amino acid sequences of proteins, and on DNA sequences of genes. Such data invariably exhibits some form of randomness, arising from its manner of collection, the location of the DNA in the genome, and the evolutionary history of the molecules. Statistical methods are needed to cope with this randomness, and to permit inferences to be made concerning questions such as the relative or absolute location of genetic or other markers on the chromosomes, and the nature of the evolution of the genes or genomes.
