Thanks to modern artificial intelligence technology, researchers at the UC Davis MIND Institute have recently succeeded in accurately predicting the risk of Autism Spectrum Disorder (ASD) on the basis of specific patterns of maternal auto-antibodies, i.e., antibodies that attack the tissues of their own hosts.
“The implications from this study are tremendous,” said lead author Professor Judy Van de Water. “It’s the first time that machine learning has been used to identify with 100% accuracy MAR ASD-specific patterns as potential biomarkers of ASD risk.”
In the study, Van de Water’s team analysed plasma samples from 450 mothers of children with autism, and 342 mothers of typically developing children. All samples were collected from mothers enrolled in the Childhood Autism Risks from Genetics and the Environment (CHARGE) study.
With the help of machine learning, the researchers were able to develop and validate a test to identify ASD-specific maternal auto-antibody patterns of reactivity to eight proteins that are abundant in the foetal brain.
More specifically, the algorithm went through roughly 10,000 individual patterns and identified the three most associated with the risk of MAR ASD, namely – CRMP1+GDA, CRMP1+CRMP2, and NSE+STIP1.
“For example, if the mother has autoantibodies to CRIMP1 and GDA (the most common pattern), her odds of having a child with autism is 31 times greater than the general population, based on this current dataset. That’s huge,” said Van de Water. “There’s very little out there that is going to give you that type of risk assessment.”
In the future, these maternal biomarkers could be used for pre-conception testing of high-risk women, as early diagnosis and intervention are typically much more effective at improving the quality of life of children with ASD.
Such tests – based on the quick and accurate ELISA (Enzyme-Linked-ImmunoSorbent Assay) platform – could allow women to find out the risk of their child having MAR SAD which, if ruled out, would mean that their overall risk of having a child with autism is 43% lower.
Van de Water is now working on animal models to develop effective interventions aimed at blocking the effects of these maternal antibodies on the foetal brain.
“This study is a big deal in terms of early risk assessment for autism, and we’re hoping that this technology will become something that will be clinically useful in the future.”
The paper was published on 22 January in the journal Molecular Psychiatry.