This past spring, Christian Schaaf sat back and watched seven-year-old Lily play in his office at the Baylor College of Medicine in Houston. She looked just like any other girl her age, he recalls, but she didn’t seek interaction or even eye contact in the way a child normally would. Instead, she communed with a corner of the room, excitably hopping and flapping her arms as if that spot held a treat too great to bear. Without peering into the file in front of him, Schaaf knew what afflicted Lily. “I’ve seen enough children that when I see someone with autism, I have a high suspicion for it,” he says.
Lily (not her real name) and her mother didn’t come to Schaaf’s office that day for a diagnosis; a psychiatrist had already detected autism after her fourth birthday. They visited Schaaf, a clinical geneticist, to search her genome using a chromosomal microarray. The technology can find duplications or deletions of small segments of DNA, known as copy-number variants (CNVs), to pinpoint the genetic aberration that might have caused the disorder. Lily’s parents hoped that a genetic diagnosis would help them better understand and treat her specific form of autism—and, ultimately, help her get the services she needs to have the best chance at adult independence.
Such genetic tests for autism have only become available in the last few years. But, owing to high demand, autism testing has expanded from research centers to private companies. In the US, six companies now offer laboratory-developed tests to doctors that specifically target the developmental disorder, searching the genome for either irregular CNVs or single-nucleotide polymorphisms (SNPs) that could explain the symptoms. And these tests aren’t cheap: a microarray costs, on average, $1,500, and that’s without the bells and whistles such as doctor visits and additional gene sequencing. Although the tests themselves aren’t therapeutic, they represent the leading edge of a deeper genetic understanding of autism that could lead to targeted therapies—a market that UK-based research publisher Global Data expects to top $5 billion in the US in 2018, according to an October report.
For the most part, the diagnosis of autism remains the domain of psychiatrists, who do so on the basis of a range of symptoms, including delayed speech, repetitive behaviors and social withdrawal. These abnormalities remain difficult to detect until a child is around four years of age or older, which is unfortunate because receiving therapy from age two can improve outcomes for youngsters with developmental disabilities. “The earlier the diagnosis, the earlier you can start some type of interventional therapy,” says Stephen Scherer, director of the Centre for Applied Genomics at Toronto’s Hospital for Sick Children.
The diagnostic holy grail is a molecular test that can pinpoint the disorder at birth to hook children into therapies straightaway. Just a few decades ago, this suggestion would have sounded ludicrous. From the 1950s through the 1970s, doctors thought autism resulted from poor parenting and social conditioning by ‘refrigerator mothers’, so called for parents supposedly being cold with their kids. “Now there has been a paradigm shift,” says Schaaf. “We think that 80–90% of what causes autism is really the genetics."
(Click here to continue reading.)
Most laboratory mice, when meeting new cagemates, will sniff the strangers thoroughly. But the mice in 
Last year, the World Health Organization released updated procedures on how best to tackle the global scourge of tuberculosis. The fourth edition of the “Treatment of tuberculosis: Guidelines” recommended, among other changes, increasing the dosage of tuberculosis medication required to treat children. But, in a sense, the new guidance provided a destination without a map: it failed to address the larger problem of how to improve the accuracy of pediatric dosing.
