Navigating the Genome

Around the world, life's genetic blueprint is 99.99 percent similar from person to person, with the greatest degree of variation found inside Africa, where DNA has been evolving for the longest period of time. But, single-letter variations, called "snips," for single-nucleotide polymorphisms (SNPs), may determine each person's disease-vulnerability and response to different medications, as well as a broad range of other traits -- from eye color and height to hair color and body type.

The relationship between such traits and single DNA-letter variations is poorly understood, though, and scientists also know little about the importance of common versus rare SNPs. An estimated 7 million common SNPs can be found within 5 percent of the entire human population, the Science paper notes. Another 4 million SNPs are far less common, turning up in only 1 percent to 5 percent of the world's entire population. Still more rare DNA variations may be found only in a single individual.

Shuffling the Genetic Cards

To investigate genetic variation, the Science authors took advantage of patterns among interconnected SNPs. A small, known section of genetic code containing an SNP can be used to predict larger, related chunks of sequence and, thus, provides a shortcut for mapping whole-genome patterns of genetic variations shared by people of European American, African American and Han Chinese American ancestry.

Key to the research, explains Barbara Jasny, senior supervisory editor for Science, is a phenomenon of genetic inheritance called "linkage disequilibrium," which is related to the distance between two variants, or alleles, of one gene in a single person. The further the distance between two gene variants, Jasny explained, the greater the odds that one variant may be recombined or shuffled out of the sequence block as DNA moves from parent to child.

Conversely, if the variations are closer together, they are more likely to remain together following DNA recombination over generations. At the same time, she said, haplotype blocks of genetic sequences containing genetic variations seem to have been preserved for thousands of years.

Genetic patterns may help point the way toward improved disease treatments. "Imagine that friends are giving you directions," Jasny suggests. "They might tell you that their house is part of a development that is three street lights south of the movie theater. Finding the street lights and the movie theater will be important for reaching your destination. Similarly, because genetic markers are not randomly associated, learning about a subset of them can help researchers find associations with disease-related genes."

'Bins' of Genetic Variation

During World War II, cryptographers could take seemingly random jumbles of numbers and convert them into crucial messages. Similarly, mapping strategies described in the Science paper seem to bring aspects of the human genomic sequence into sharper focus. Hinds, Cox and colleagues used a special mathematical algorithm, based on the principle of linkage disequilibrium, to characterize the structure of genetic diversity by distributing different SNPs within interconnected bins of genetic code spanning the whole genome. Genomic data also were assigned to haplotype maps, which were connected with bin patterns.

Before they could identify bin or haplotype patterns, though, researchers first had to investigate 2,384,494 SNPs believed to be common across diverse populations. Of these genetic differences, some 69 percent were identified by sequencing DNA samples from 24 people, then filling in holes in the resulting information using SNPs from public databases.

Next, the researchers analyzed the genetic blueprint of 71 people who were not related to each other or to the 24 people studied in the initial analysis of SNPs. The group of 71 people included 24 European Americans, 23 African Americans and 24 Han Chinese American individuals whose DNA had been archived within the Coriell Cell Repositories' Human Variation Collection. Of the 2.4 million SNPs originally described by the researchers, some 1,586,383 were found in both variant forms, or alleles, among the 71 DNA samples.

Because 157,000 of the SNPs and nine individuals from their study had previously been analyzed by the HapMap Project, which assessed 1 million SNPs in 270 people, the Science authors were able to compare the two sets of data. In more than 99 percent of all assessments, results were identical in both studies, suggesting that "both data sets are of exceptionally high accuracy," Altshuler notes.

Comparisons with data from a Seattle-based study by Debbie Nickerson and colleagues offers further confirmation of accuracy and indicates a subset of SNPs where further analyses are necessary.

Most of the resulting 1.58 million SNPs were common to all three human populations, researchers report. Some 94 percent of the SNPs, for example, were found in both variant forms among African Americans. Among European Americans, 81 percent of all the SNPs showed up in both forms, and 74 percent of SNPs were found in both forms in DNA from individuals of Han Chinese American ancestry.

'Spectacular' Ressource for Geneticists

How useful is the new mapping effort? In particular, if some diseases are caused by very rare genetic variations, will a map of common variations prove truly practical? "The biological insights should be of tremendous value," Atshuler said, while also acknowledging the possibility of exceedingly rare disease-related variations. "In addition to the potential utility for disease research, such data are a spectacular resource for population and evolutionary geneticists," he concludes.

The SNPs analyzed in this study represent only a fraction of the more than 10 million common SNPs expected to exist in the human genome. But, Hinds and colleagues demonstrate that with just this relatively small number of "street lights," they should be able to find their way to most of the common variants in the human genome, Jasny says. "As few as 300,000 or 500,000 SNPs could give us most of the information," Hinds adds.

Although this research probably won't allow doctors to identify individual risk factors, scientists report, "knowledge of a large fraction of all the major genetic risk factors contributing to a treatment response or common disease could have immediate utility, allowing existing treatment options to be matched to individual patients without requiring additional knowledge of the mechanisms by which the genetic differences lead to different outcomes."

The Science paper, "Whole Genome Patterns of Common DNA Variation in Three Diverse Human Populations," was authored by David A. Hinds, Laura L. Stuve, Geoffrey B. Nilsen, Dennis G. Ballinger, Kelly A. Frazer, and David R. Cox of Perlegen Sciences, Inc., Mountain View, California; with Eran Halperin of the International Computer Science Institute of Berkeley, California; and Eleazar Eskin of the University of California at San Diego. The research makes use of previously reported data from the International HapMap Project.