||30 times to achieve geometric amplification of the region. With the whole genome sequence in hand any region can be amplified for study. So it is now almost trivial to compare one region in 1000 people or 100 regions in fifty people. In fact it is this technology that helps to set the accuracy limits for a reference DNA sequence â€" in the extreme it only really needs to be good enough to be able to reliably predict primers that will work. Sequence Variation Sibs 1/2000 Unrelated People 1/500 Chimps 1/60 Mouse 1/30 Spinach 2/3 We have a general idea about variation of DNA sequence among human populations, but specific details are lacking. Identical twins have virtually identical sequences. Siblings differ by 1/2000, a statistic which provides the basis for family studies to follow the inheritance of a disease trait and map its chromosomal location. Unrelated humans differ by one base in every 500, which translates to 6 million differences in the 3 billion-base genome. It is these sequence differences that account for all the inherited phenotypic differences between us and that are the basis for DNA forensics. (Since 1/500 bases are different between any two people, there may be little point in having an accuracy much greater than this). How this variation is spread over the genome is not clear. Large-scale sequence comparisons among people will be very revealing. Sequence comparisons will be very powerful for probing complex genetic diseases that involve many genes. Qene Function Once the encyclopedia of human DNA is complete, it will have some 100,000 genes embedded in the 3 billion-base genome. How do we find the genes and their functions? A powerful approach to finding genes is by comparison of sequences among different species. Functional sequences, genes, are highly conserved through evolution while other regions have diverged considerably. Comparisons often also assign functions.