The Icelandic Cancer Project involves a comprehensive linkage disequilibrium search for cancer predisposing genes, with a key component of a whole genome scan.  All living cancer patients, their relatives, and controls are invited to participate, and the great majority does.  It draws on the model features of the Icelandic population isolate of good genealogical records, standardized health care, and reliable medical information, in addition to the advantage of greatly increased intervals of linkage disequilibrium.

  We outline in this talk the promises and execution of the project, from the view of a computer scientist, and discuss some of the bioinformatics challenges.  In particular, we describe in some detail our algorithms and prototype system for selecting markers and primers for multiplexing genotyping experiements.