Local pressures such as climate or diet that affect natural selection are only partly responsible for differences in the genetic makeup of human populations, a new study finds. Migrations, the expansions and contractions of populations and the vagaries of genetic chance also play a role.

“It’s easy to imagine that natural selection could have led to human populations strongly diverging from each other as they adapted to the different environments they encountered as they spread around the world,” said Graham Coop, an assistant professor of evolution and ecology, and lead author of the paper describing the study. “But what excites me about this work is that we’re showing that a population’s history and migrations between groups have had a strong role in limiting and shaping how populations adapted to their environments.”

The study, led by Jonathan Pritchard, a Howard Hughes Medical Institute investigator at the University of Chicago, was published June 5 in the online journal PLoS Genetics.

Variants do not explain everything

Natural selection occurs when a particular genetic difference — known as a variant — gives an individual a greater opportunity to have children and pass on his or her genes to future generations.

A good example are the variants responsible for light skin color. As modern humans moved out of Africa and spread into northern latitudes, dark skin became a disadvantage, possibly because it blocked too much of the sunlight necessary for synthesis of vitamin D for healthy bones. Genetic variants for light skin, then, conferred a small survival edge, and today those variants are common in people of European and northern Asian ancestry.

Yet genetic variants do not always help populations adapt to their environments, Pritchard said. For example, if a small population undergoes a rapid expansion — because it has entered new territory, perhaps, or developed a technology that supports larger numbers — some of the genetic variants carried by that population can increase rapidly in number, even if they do not provide a reproductive advantage. Also, the pool of genes within a population tends to fluctuate due to chance events and random differences in the number of children people have and the particular genes they pass on to their offspring.

Human Genome Diversity Project

Thus, one of the fundamental questions facing human geneticists is: Is it possible to determine which genetic variants have increased because of selection and which have increased because of population changes or genetic chance?

Pritchard, Coop and their collaborators decided to tackle this question when new genetic data became available last year from the Human Genome Diversity Project at Stanford University. These data provided a much more in-depth sampling of worldwide genetic differences than had previously been available, allowing the research team to carry out a new and more rigorous test for selection.

To determine whether a variant’s frequency resulted from natural selection, the team compared the distribution of variants in two parts of the genome: in regions that affect the structure and regulation of proteins, and in neutral regions that have no effect on proteins. Because neutral regions are less likely to be affected by natural selection, they should reflect the demographic history of populations, the team reasoned. In contrast, regions that have been influenced by selection — the protein-regulating regions — should show different patterns of distribution.

The analysis immediately identified known examples of selection, including those involved in determining skin pigmentation, resistance to pathogens, and the ability to digest milk as an adult, a trait that arose in Europe, the Middle East and Africa following the domestication of dairy animals.

Interestingly, it also revealed several genes of unknown function that appear to have been under strong selective pressures. “We’re keen to learn what these genes are and how they work,” Coop said.

Historical and demographic factors

Yet the team also found that many genetic signals that other researchers have attributed to selection may actually have been created by historical and demographic factors. When the group compared closely related populations — those that recently came from the same ancestral population or those that have exchanged many migrants throughout history — it found few large genetic differences. If these populations’ environments were exerting strong selective pressure, such differences should have been more apparent.

Even with genetic variants where evidence of selection is strong, such as those for skin color, Pritchard noted, the movements of populations have powerfully influenced current patterns of variation. “A handful of selective signals are clear,” he said, “but it’s hard to be confident about individual cases beyond the top 10 or so that we understand well right now.”

Senior public information representative Liese Greensfelder and the Howard Hughes Medical Institute contributed to this report.