| Despite over 30 years of effort, it remains unclear as to how duplicate genes are retained in a genome. To pursue this issue, I studied factors that may affect gene duplicability, which is defined as 1 - P, where P is the proportion of unduplicated genes in the genome. First, I considered the dosage balance hypothesis, which postulates that duplication of a gene encoding a subunit of a protein complex may cause dosage unbalance among the subunits of the complex and thus is less likely to be retained than duplication of a gene encoding a monomer. Using data from yeast and human, I found evidence for this hypothesis, i.e., P increases with the number of subunits in the protein complex. This increases is initially slow but becomes rapidly when the complex reaches a certain size. Second, I considered developmental constraint. Using large-scale gene expression data of fly development, I found that the average duplicability for genes involved in embryonic development is indeed lower than that for genes involved in larval development but not significantly lower than that of genes involved in later stages of development. Moreover, in both flies and nematodes genes with multiple expression peaks do not show a lower duplicability than genes with a single expression peak. Therefore, the effect of developmental constraint on gene duplicability may be weak. In addition, I studied the relationship between sequence divergence and expression divergence of duplicate genes. I developed a new measure for expression conservation. Application of this measure to large scale expression data in human and mouse revealed that expression breadth (number of tissues expressed) is an important determinant for evolutionary conservation of tissue expression. Further, I found a fast rate of divergence in tissue expression between duplicate genes. Finally, in a study of the relationship between protein dispensability and evolutionary rate, I found only a very weak correlation and that this correlation is largely due to the existence of duplicate genes. |
