With symptoms that can range from missed social cues to severe linguistic and cognitive impairments, autism spectrum disorder (ASD) has proved a complex condition for geneticists to get their heads around.
Now the largest study to date based on the whole genome sequences of siblings with ASD, together with their non-autistic parents, is throwing a genetic spotlight on those complexities and yielding some surprises.
Among them: In only one third of the cases where the autism of one sibling with ASD was linked to a genetic variant did the other sibling with autism share the same variant.
At face value, such a result might seem to defy common sense. Autism is thought to affect about 1 in 68 children, which means the odds of two siblings having the disorder for entirely unrelated reasons should be very low.
One possible explanation is some of the variants the study looked at will prove in time not to be implicated in autism. Or there could be other still-hidden inherited factors that the siblings share that may increase the likelihood of ASD in some way. "Then, if they have another mutation, they're pushed across the autism threshold," says Stephen Scherer, director of the Centre for Applied Genomics at the Hospital for Sick Children in Toronto who led the study.
The possibility of different autism-linked genes coming into play among siblings with ASD matters for parents trying to choose the right treatments for their children. Even when outward symptoms are similar, the underlying genetics factors may call for completely different interventions.
And for families wondering whether they have more than one child with ASD when one child has a disorder that has been linked to a genetic variant, it is probably not enough to simply test to see whether other children in the family share that particular variant.
"You need to look at the whole genomes," Dr. Scherer said. "Two thirds of the time it's going to be something different."
The study, which includes whole genomes from 85 families – 340 individuals in total – represents a significant step toward a larger initiative supported by the foundation Autism Speaks that ultimately aims to sequence 10,000 genomes. In conjunction with the study's publication Monday in the journal Nature Medicine, the 340 genomes are among 1,000 that are now being made available to outside researchers via a platform developed by Google.
"It's a great resource and an important one for the field," said Joris Veltman, a professor of translational genomics at Maastricht University in the Netherlands who was not involved in the SickKids study.
Estimates of the total number of possible genetic variations involved in ASD ranged from about 400 to more than 800. To date, only about 100 genes have been linked to the disorder with a relatively high degree of confidence.
Most of those have been identified through searches that concentrate on the parts of the human genome – roughly one per cent – that are known to be involved in biological function. Such an approach is more cost effective than whole genome sequencing, "but we miss a lot," said Michael Ronemus, a researcher who specializes in the genetics of autism at Cold Spring Harbor Laboratory in New York.
Dr. Ronemus predicted the SickKids study will herald many others that take a similar approach as the field rapidly moves toward whole genome sequencing to capture the full portrait of an individual's genetic variation and see how that may play into conditions like ASD. He is involved with another research consortium that is working in precisely that direction.
Researchers are still trying to ascertain how the various genetic variations that do arise in connection with ASD may be having an effect. But while the precise mechanisms remain to be discovered, the arrival of whole genome sequencing as a diagnostic tool for families is already motivator for those participating in the SickKids effort.
"It's to find out why," said Valerie South, a Toronto mother who has two of three sons with autism but whose oldest boy, now 21, never showed any symptoms of the disorder.
After her second was found to have autism, it would be six more years before Ms. South decided to have one more child. At the time she recalls being told that the likelihood of the same thing happening again was like lightning striking twice.
Her third son, now 14, also has ASD with a diagnosis that, on paper, closely mirrors that of his older sibling. But in many ways the two are entirely unalike, she says.
While her middle son tends to sit, her youngest darts around "like a bullet." He also interacts with an iPad and responds to signing, none of which interests his older brother.
"Their autisms are expressed differently," Ms. South said of her two sons with ASD. "I can see the differences in the two boys and it is marked."
Later this year Ms. South and her family will add their genomes to the growing database the Dr. Scherer is building. It may well show that the specific genetic underpinnings of her two sons' ASD are not identical, though that remains to be seen.
"What's important is finding out answers," she said. "We all need to come together to do this."
The database includes not only whole genomes sequences but details about how ASD manifests in those people who have been sequenced.
"You can correlate what their life is like with what their molecules are like," said David Glazer, engineering director for Google Genomics in Mountainview, Calif., which is hosting the database.
Once researchers have been approved for access they can analyze the data themselves using the tech company's cloud platform to look for hidden patterns in the DNA and follow their own hunches, he said.
"The vision is really to take this information and make it as useful to as many people in as many places as possible."