A database system for constructing, integrating, and displaying physical maps of chromosome 19

Efforts are underway at numerous sites around the world to construct physical maps of all human chromosomes. These maps will enable researchers to locate, characterize, and eventually understand the genes that control human structure and function. The volume and complexity of the data already generated requires a sophisticated array of computational support to collect, store, analyze, integrate and display data in biologically meaningful ways. The Human Genome Center at Livermore has spent the past 6 years constructing a database system to support its physical mapping efforts on human chromosome 19. Our computational support team is composed of experienced computer professionals who share a common pragmatic primary goal of rapidly supplying tools that meet the ever-changing needs of the biologists. In this paper, we concentrate on the design issues, tradeoffs, and consequences from the point of view of building a complex database system to support leading-edge genomic research. We introduce the topic of physical mapping, discuss the key design issues involved in our database, and discuss the use of this data by our major tools (DNA fingerprint analysis and overlap computation, contig assembly, map integration and database browsing). Given the advantage of hindsight, we discuss what worked, what didn't, and where to go from here. As early pioneers in this field, we hope that our experience may prove useful to others who are now beginning to design and construct similar systems.<
>