The study comes from the Simons Foundation, a private foundation based in New York City, and to derive at the DNA-autism connection the researchers harnessed the power of artificial intelligence. The application of AI found that mutations in so-called ‘junk’ DNA can lead to autism.
Autism spectrum disorder is a neurodevelopmental disorder that affects social interaction, communication, interests and behavior. Indications that a child is on the autism spectrum tend to come from parental observations. Here signs are noticed during the first two or three years of their child’s life.
In terms of some forms if autism, a genetic linked has been established; for example, with mutations as with in a gene called SLC7A5. The role of SLC7A5 is to transport a specific type of amino acid called branched-chain amino acids (BCAA) into the brain. There are other potential causes as well, such a connection with the human microbiome of the gut.
READ MORE: Virtual reality headsets help police with autistic people
The new study’s findings functionally link specific mutations to autism and indicates a connected between non-inherited, non-coding mutations and what is a complex human disorder. The scientists deployed machine learning to assess the whole genomes of 1,790 individuals identified autism. The data was compared to the unaffected parents and siblings of the test subjects. With each of the individuals tested, there had been no family history of autism. This led to the hypothesis that the genetic cause of their autism was connected to spontaneous mutations rather than being the consequence of inherited mutations.
It was found that mutations in protein-coding regions account for at most 30 percent of autism cases in individuals without a family history of autism. The results further suggested that mutations which affected gene expression in the brain and genes were the same as those responsible for neuron migration and development
Further assessment of the data revealed that genetic mutations in parts of the genome do not encode proteins (areas of so-called ‘junk’ DNA). The analysis showed how the number of autism cases connected to the non-coding mutations was equivalent to the number of cases linked to protein-coding mutations that disable gene function.
The new research has been published in the journal Nature Genetics. The research paper is titled “Whole-genome deep-learning analysis identifies contribution of noncoding mutations to autism risk.”
