Teen Inventor Designs Noninvasive Allergy Screen Using Genetics and Machine Learning
Seventeen-year-old Ayush Alag is one of 40 finalists in this year’s Regeneron Science Talent Search
One of Ayush Alag’s earliest memories is of biting into a chocolate bar with cashew nuts and suddenly feeling his throat get itchy.
For most of his childhood, the Santa Clara, California resident avoided eating anything with cashews and other nuts that caused irritation as best as he could. By his middle school years, he and his parents wanted to know for sure: did he have a serious food allergy, like 32 million other Americans, or was it just a food sensitivity? They sought the help of an allergist, Joseph Hernandez of Stanford University.
Hernandez told them that the difference between an allergy and a food sensitivity is huge. In the case of food sensitivity, a person can slowly introduce the reaction-triggering food back into their diet little by little to build immunity. If you’re allergic, however, doing so could result in death.
Hernandez recommended that Alag first take a blood test and a skin test, both typical measures in determining allergies, but those results were inconclusive, which was frustrating for Alag when he knew that eating certain foods made him ill. Blood and skin tests are hyper-sensitive and produce false positives nearly 50 to 60 percent of the time, according to the nonprofit Food Allergy Research & Education based in McLean, Virginia. The only way to truly know whether or not he was allergic was to do an oral food challenge, an experience that can be a stressful and occasionally traumatizing for the patient. During an oral food challenge, the patient eats three small amounts of the suspected allergen over the course of an hour under the supervision of a doctor and nurse. Then the patient is observed for four hours following the last dose to see if symptoms occur.
Not only is the process time-consuming for a medical provider, it is also risky, especially for children. If a severe reaction occurs, the child will need to be transported to an emergency room as soon as possible. Food allergies affect approximately eight percent of children, and in decades of oral food challenges being an industry standard there has been just one reported fatality. In 2017, a three-year-old died during a routine challenge, startling the allergy research community.
Thankfully, Alag learned that he did not have a serious allergy after his oral food challenge, but rather a simple food sensitivity, and he was able to make a plan to reintroduce food that used to cause him irritation back into his diet. Now, he says, he can even order cashew Pad Thai without any trouble at all.
“As someone who has been through this whole process and knows what a life changing difference it can be to correctly be classified as sensitized and not allergic, it motivated me to research if there is a way that I can diagnose food allergies that is both safe and accurate,” Alag explains.
After his experience, Alag, then 14, decided to look for solutions on his own. At the time, he had been learning about algorithms in school. At home, he was reading previous studies that identified specific genes associated with certain allergies. Driven by his interest in computer science, he designed an algorithm that successfully flags genetic markers linked to food allergies using publicly available data sets. (He focused on 18 genetic markers that were relevant to what he wanted to achieve with allergy testing.) In theory, all you’d need to do is give a blood sample and his test would indicate whether you have allergies to a certain substance or not. Genetic testing is the general direction the field is headed, but the major obstacle that stands in the way is a need for a bigger sample size.
Now, at 17, he runs his own company called Allergezy. (His childhood allergist, Hernandez, is now a clinical partner with Allergezy.) He is also a top 40 finalist for the Regeneron Science Talent Search for high school seniors, which concludes tonight with a gala, where winners will be announced at the National Building Museum in Washington, D.C. Regeneron will distribute a total of $1.8 million with a top prize of $250,000 for the first place award, according to the Regeneron website.
“These kids are working on projects that could truly change lives around the world. In Ayush’s case, he could help improve diagnosis for the increasing number of people suffering from food allergies,” says George Yancopoulos, president and chief scientific officer of Regeneron (who was also once a finalist in the competition), in a statement via email.
Alag’s technology is far from market-ready; it still needs a lot of refining, which will require sequencing more DNA samples, and testing to prove that it is reliable and accurate. Allergezy recently received a $10,000 grant from the Silicon Valley genetic research startup Illumina to expand the datasets, obtain more blood samples and do more genetic sequencing. (All of that and much more needs to happen before they approach the FDA approval process.) He is one of nine Regeneron finalists who have applied for a patent, notes Maya Ajmera, president and CEO of the Society for Science & the Public, in a statement via email.
“Ayush’s machine-learning algorithm, which can safely diagnose a food allergy from a patient’s blood profile, has the potential to make a real impact on people’s lives,” says Ajmera.
And that’s exactly what Alag hopes to do some day. He explains how much of a relief it is for him not to live with the constant fear that he might have a reaction to the food in front of him, and he just wants others to feel that same sense of freedom.
“It’s a lot easier now for me and my entire family, and it’s a much better lifestyle change,” Alag says.
But, one thing at a time. First, he has to figure out where he’s going to college.