Diverse datasets have become key building blocks of translational biomedical research. Data types captured and referenced by sophisticated research studies include high throughput genomic and proteomic data, laboratory data, data from imagery, and outcome data. In this paper, we present the application of an XML-based data management system to support integration of data from disparate data sources and large datasets. This system facilitates management of XML schemas and on-demand creation and management of XML databases that conform to these schemas. We illustrate the use of this system in an application for genotype-phenotype correlation analyses. This application implements a method of phenotype-genotype correlation based on phylogenetic optimization of large datasets of mouse SNPs and phenotypic data. The application workflow requires the management and integration of genomic information and phenotypic data from external data repositories and from the results of phenotype-genotype correlation analyses. Our implementation supports the process of carrying out a complex workflow that includes large-scale phylogenetic tree optimizations and application of Maddison's concentrated changes test to large phylogenetic tree datasets. The data management system also allows collaborators to share data in a uniform way and supports complex queries that target datasets.