A substantial fraction of almost every genome sequenced consists of genes that have no known function and that do not have sequence similarity to any genes with known function. How would your approach change if the genes of unknown function were in the human genome?