Abstract

Much scientific data can be characterized by properties like complexity, large volume, low update frequency, and indefinite retention, which brings up some different issues than those found in conventional business environments. There are a number of influences that should guide the development of data models in bioinformatics. These range from experience of the scientific database community across a range of disciplines, the current best practice in bioinformatics system, available data models and schemas, the impact of emerging standards, and the trend towards ontologies. Influenced by them, I have developed a conceptual object data model for the sequencing quality control and assembly pipeline for the genomics project creating a cDNA library for Aspergillus niger and implemented it using the relational database MySQL. This experience is summarized in a set of Guidelines for data modeling in bioinformatics.