An international team of researchers has published a study showing that more than 10 percent of human genes vary in the number of copies of DNA sequences they contain — a finding that contradicts previous assumptions that the DNA of any two humans is 99.9 percent similar.
The study was published in last week’s Nature and was conducted by researchers at Brigham and Women’s Hospital, Harvard Medical School, the Hospital for Sick Children, the Wellcome Trust Sanger Institute, the University of Tokyo, and Affymetrix [Redon R, et al. Global variation in copy number in the human genome. Nature. 2006 Nov 23;444(7118):444-54].
The discovery indicates that copy number variation could play a larger role in genetic disease than previously thought, with broad implications in disease association studies, genetic diagnostic testing, and cancer research. Indeed, paper co-authors are now arguing that a CNV detection step should be mandatory in all disease-association studies.
According to Lars Feuk, a research associate at the Hospital for Sick Children in Toronto and a co-author on the paper, “everyone that is doing a disease-association study now should incorporate a copy number variation detection step in their study design.”
“We have been arguing for that for awhile now, and of course, with the amount of variation that we found out, we know that it is important. There also have been several diseases now where copy number variation has shown to be a susceptibility factor for disease. The CCL3 gene in HIV susceptibility is one example,” Feuk said.
He added that while the addition of a CNV step will undoubtedly create more work for researchers, it should not be avoided out of laziness.
Feuk and his co-authors used two methods — Affymetrix human mapping 500K arrays and whole-genome tilepath arrays with 26,574 large-insert clones — to study DNA samples from 270 participants in the International HapMap project. They found around 2,900 genes — more than 10 percent of the genes in the human genome – that vary in copy number from the predicted two inherited copies. Previous research only looked at single-base-pair changes among this study group.
The study team found an average of 70 copy number variations averaging 250,000 nucleotides in size in each DNA sample. Overall, the researchers identified 1,447 different copy number variations covering around 12 percent of the human genome and between 6 and 19 percent of any given chromosome.
Feuk said that, while Affy played a significant role in the project, researchers have a variety of options available to them in order to include a CNV step in a disease-association study.
“Which platform you use is an open question. There are several different platforms that can be used. Of course if you wanted to do a SNP association study, you could use another type of platform to look at copy number variation,” he explained.
Feuk said that both Agilent and NimbleGen sell comparative genomic hybridization products that could be used in conjunction with SNP analysis tools for detecting copy number variation. He also said that Affy’s mapping arrays worked well for the purpose of the study because “you can get both CNV data and SNP data out of the same platform.”
“You don’t get as much copy number variation data as you would from a CGH array that is specifically made for the purpose of finding of copy number variation, but you still will find a lot,” he said.
Feuk added that by the time some of the large disease-association studies get around to including a CNV detection step, the technology will most likely have improved to provide researchers with ways to do both SNP and CNV detection. He said that all array companies should be looking at a way to offer CNV detection capabilities to their clients.
This article originally appeared in this week’s issue of BioArray News, a Pharmacogenomics Reporter sister publication.