NEW YORK (GenomeWeb) – A University of Washington-led team has started characterizing copy number variants that differ from one human population to the next, providing new clues to population migrations, relationships, and ancestry patterns.
As reported online today in Science, through CNV profiling on individuals from more than 100 worldwide human populations, the researchers saw that duplications often show more between-population differentiation than deletions. Their global CNV analysis uncovered a few large stretches of sequence that have disappeared in specific human lineages. But the distribution of the deletions, and population details they provided, differed from those offered by duplications.
"We find that duplications and deletions exhibit fundamentally different population-genetic properties," senior author Evan Eichler, a genome sciences researcher at the University of Washington, and colleagues wrote.
"Duplications are subjected to weaker selective constraint and are four times more likely to affect genes than deletions," they explained, "indicating that they provide a larger target for adaptive selection."
The team found that African populations had higher ancestral human sequence levels since they have undergone fewer population bottlenecks than their counterparts from other populations. Consequently, duplications and deletions appeared to make up a proportionally larger piece of the genomic variation found in individuals from non-African populations, though overall CNV numbers were comparable across populations.
Still, the study's authors emphasized that the "biological significant of this difference should be interpreted cautiously and will require association studies to determine its relevance to disease and other phenotypic differences."
Although a slew of human diversity studies have already been done, the researchers explained, most have focused on SNP patterns or CNV profiles produced from various types of array data.
For the current analysis, the team used genome sequencing to profile CNVs across human populations, generating an average of 41-fold genome coverage per person for 236 individuals from 125 populations with Illumina's HiSeq 2000. The group also considered existing genome sequences for Neanderthal, Denisovan, and ancient human individuals.
By analyzing the sequences with a digital comparative genomic hybridization read-depth approach for uncovering copy number alterations, the researchers detected almost 14,500 autosomal CNVs and another 545 CNVs involving sequences on the X chromosome — patterns they verified with SNP- and array comparative genomic hybridization microarrays.
More than 80 percent of the CNVs were smaller than 25,000 bases apiece, they reported, though those found within segmental duplications stretch out further, on average. All told, these CNVs contributed to variation found across roughly 7 percent of the average human genome.
But some types of copy number changes were far more common than others. Whereas almost two-thirds of the CNVs the team identified were duplications, just one-third or so were deletions. Likewise, the analysis revealed almost twice as many population-stratified duplications compared to deletions differing between populations.
The researchers noted that African populations clustered separately from other populations based on either duplication or deletion data. On the other hand, duplication patterns most clearly differentiated Oceanic populations such as those in Papua New Guinea, Australia, and Bougainville.
As found in prior SNP studies, populations from Papua New Guinea also shared higher-than-usual CNV-based genetic similarities with the ancient hominin group known as the Denisovans.
Most notably, the team uncovered a composite segmental duplication on chromosome 16 that was apparently passed from Denisovans into the ancestors of individuals in Papua New Guinea and Bougainville. That duplication is now nearly fixed in those populations, but didn't show up in humans from other parts of the world.