Model system: Researchers have built a map of nearly 2,000 interactions among 418 genes in yeast cells, which are easy to manipulate genetically.

The pattern of interactions among different genes in yeast cells changes in response to disease-like conditions, in this case a DNA-damaging agent, according to a study published 3 December in Science. Mapping epistasis — how various cellular factors work together — is key to understanding complex disorders, such as autism.

Studies looking for mutations linked to autism have identified many gene candidates, but few of these are likely to be sufficient to cause the disorder on their own. Researchers instead say that autism is probably the result of multiple genetic variants working in combination.

Traditional genetic methods cannot fully map these interactions, says Trey Ideker, chief of genetics at the University of California, San Diego. “Even after sequencing everyone on the planet, we’re still not going to have enough power to find these combinations,” he says. The solution, says Ideker, is to create a model network that can pinpoint gene combinations of interest.

The researchers created a model in yeast cells by combining 418 mutations into 80,000 pairwise combinations. Yeast colonies that grow better or worse with one of these combinations when compared with either mutation alone indicate an interaction between the two genes. The resulting network map identified 1,905 of these interactions.

When the researchers perturb the system by introducing a chemical that damages DNA, the interactions change dramatically. Of the 2,297 interactions under this condition, 70 percent are new.

Ideker says this kind of response is probably not unique to the DNA damage pathway. He says different diseases may each lead to a cellular response with a unique set of interactions. Similar diseases could result in the same pattern of genetic interactions, however, which would “tell us something fundamental about the disease and how to treat it.”

Ideker’s team is looking at changes in the yeast network under a variety of different conditions to confirm this theory. The next major challenge, he says, is to develop the technology needed to test this in human cells. “We expect a similar story to emerge in humans.”