Investigating CpG islands in mammalian genomes

描述（由申请人提供）：

DESCRIPTION (provided by applicant): CpG islands, clusters of CpG dinucleotides in GC-rich regions, are often located in the 5' end of genes and are considered gene markers in mammalian genomes. Experiments have shown that methylation of promoter CpG islands plays an important role in gene silencing, genomic imprinting, X-chromosome inactivation, and carcinogenesis. Although many studies have been performed to identify CpG islands, to estimate the CpG mutation rates, and to examine the methylation modulation on disease genes, little is known about the distribution of CpG islands in the specific genomic regions across mammalian genomes or the pattern of genetic variation in CpG islands or in promoter-associated CpG islands. In addition, how CpG islands in mammalian genomes evolved and which kind of genes tend to gain or lose CpG islands remains largely unknown to us. The recent releases of several mammalian genomes, the discovery of millions of single nucleotide polymorphisms (SNPs) in these genomes, and large-scale methylation status data in the human genome provide us with an unprecedented opportunity to examine these issues in details. In this project, we will first perform a comprehensive survey of the distribution of CpG islands in five mammalian genomes (human, chimpanzee, mouse, rat, and dog) and their genomic regions. We will compare the features of CpG islands in the housekeeping genes with those in tissue-specific genes and compare the distribution and features of CpG islands in different categories of Gene Ontology (GO). Second, we will examine mutation patterns in CpG islands in four genomes (human, chimpanzee, mouse, and dog), each of which has more than 1 million SNPs. Third, we will investigate the mechanisms of CpG island evolution in mammalian genomes and to infer its effect on gene function. One approach is to examine the mutation pattern (e.g., CpG??TpG/CpA) and sequence features in each segment of CpG islands. Using the HapMap SNPs with their allele frequencies, we will test the possible natural selection in CpG islands. Finally, we will test the hypothesis that there is a relationship between CpG island loss and methylation status in CpG islands. Our comparative genomics and biomedical informatics analyses will provide a detailed view of the CpG island distribution in mammalian genomes, the mechanisms of CpG island evolution, and the relationship of the mutation pattern in CpG islands with functional categories of genes, especially cancer-related genes.