Noboru Hiroi, PhD, professor in the departments of Pharmacology, Cellular and Integrative Physiology, and Cell Systems and Anatomy, is studying mice that have this deletion. The team is harnessing the power of machine learning to understand which vocalizations of the newborn mouse pups are most predictive of social abnormalities one month later when the pups reach puberty.
“It is essential to identify those very early signs that can predict what is to come, because if we can translate what we discover in mouse pups to human infants and apply therapeutic options earlier, their outcome will be better,” Hiroi said.
When mouse pups are separated from their mothers, they emit ultrasonic vocalizations in a certain order. Mouse mothers respond to the cries and take care of their offspring.
The cries are abnormal in mice with the 16p11.2 deletion.
“Pups that carry this genetic variation cannot form the correct sequence,” Hiroi said. “We want to know whether those abnormal sequences or combinations of call types can predict what is to come one month later in their social behaviors.”
Machine learning will enable the team to develop a precise diagnostic algorithm for autism spectrum disorder.
“Once we can do this with the mouse vocalizations, we can apply the same algorithm to the cries of human babies,” Hiroi said.
Infants at risk of the disorder who are identified in this way can be given desensitization therapy so that they don’t overreact to certain cues they don’t like, he said. And behavioral therapies can be applied to help babies learn how to cope in social situations.
The research is described in the journal Molecular Psychiatry.