3 – Managing Pedigree and Genotype Data

Model-based or model-free analyses of genetic linkage data are popular ways of mapping disease genes using extended families or affected sib-pairs, and the statistical ramifications of this work are well developed. However, the management of information to facilitate the process, particularly with regard to the transformation of data formats and support for statistical analysis, remain a bottleneck in any mapping project. This chapter discusses the domain of human pedigree analysis, with particular reference to the management of genotypes and phenotypes, and considers how experimental data generated by this kind of work can be managed effectively. The discussion is based around well-established data modeling techniques such as Entity Relationship Attribute (ERA) analysis, which is an appropriate design methodology for relational database systems, and provides example code in Structured Query Language (SQL) form. The intention of this discussion is to provide a basis for the construction of a real working system, which can be implemented in one of a number of industrial database technologies, such as Oracle, Sybase, or Informix, as well as more leading-edge systems, such as ObjectStore.