Genome Informatics: Requirements and Challenges

ABSTRACT Informatics of some kind will play a role in every aspect of the Human Genome Project (HGP): data acquisition, data analysis, data exchange, data publication and, data visualization. What are the real requirements and challenges? The primary requirement is clear thinking and the main challenge is design. If good design is lacking, the price will be failure of genome informatics and ultimately failure of the genome project itself. We need good designs to deliver the tools necessary for acquiring and analyzing DNA sequences. As these tools become more efficient, we will need new tools for comparative genomic analyses. To make the tools work, we will need to address and solve nomenclature issues that are essential, if also tedious. We must devise systems that will scale gracefully with the increasing flow of data. We must be able to move data easily from one system to another, with no loss of content. As scientists, we will have failed in our responsibility to share results, should repeating experiments ever become preferable to searching the literature. Our databases must become a new kind of scientific literature and we must develop ways to make electronic data publishing as routine as traditional journal publishing. Ultimately, we must build systems so advanced that they are virtually invisible. In summary, the HGP can be considered the most ambitious, most audacious information-management project ever undertaken. In the HGP, computers will not merely serve as tools for cataloging existing knowledge. Rather, they will serve as instruments, helping to create new knowledge by changing the way we see the biological world. Computers will allow us to see genomes, just as radio telescopes let us see quasars and electron microscopes let us see viruses.