The day when you can sequence your baby's genome before it is born might not be too far away. Researchers have reconstructed the genome of a fetus without touching it. Instead, they used both parents' genomes and free-floating fetal DNA, which circulates in the mother's blood.
Several companies already offer non-invasive prenatal tests for diseases caused by having extra copies of chromosomes, such as Down's syndrome, which is also known as trisomy 21 because it involves an extra copy of chromosome 21.
These tests are based on examining the fetal DNA that crosses the placenta, comprising up to 13 per cent of the free-floating DNA in the mother's blood plasma. The ideal screen would detect every genetic disease, some of which are caused by difficult-to-detect mutations.
"If the genome is a book, and a trisomy is an extra chapter, we want to find every typo," says Jay Shendure of the University of Washington in Seattle.
In 2010, researchers used parental genomes and fetal DNA from the mother's blood to map disease-associated mutations in a fetus's genome. Although the method could be used to infer the fetal genotype, it involved sampling the placental tissue, an invasive process that may slightly raise the risk of miscarriage.
In the latest research, Shendure's group sequenced both parents' genomes, using blood from the 18.5-week-pregnant mother and saliva from the father. They then used a technique known as deep sequencing to analyse dozens of samples of fetal DNA from the mother's blood. That left them with an assortment of genetic material.
After working out which parts of the fetal DNA came from the mother and father, it was possible to see which bits were unique to the fetus, and use the information to piece together its genome. The team sequenced the genome again after birth to assess the accuracy of their fetal genome technique. Overall, the genome they pieced together was 98 per cent accurate. The fetus had 44 new mutations of its own, of which the team managed to spot 39.
Shendure says the technology can detect more than 3000 diseases that are caused by mutations in a single gene. However, fetal sequencing faces the same problems as any other genome sequence: researchers are limited in their ability to tell whether any given mutation will actually cause a specific disease. Plus, the rate of falsely predicting a child's chance of having a disease is still far too high to be useful in the clinic, he says.
It may be at least five years before the test reaches the clinic, Shendure says.