Identifying autism blood biomarkers with machine learning

The UT Southwestern team has used machine learning tools to analyse hundreds of proteins that has led to the identification of nine serum proteins that predict the disorder. The researchers hope this will help develop more effective therapies for ASD sooner. The study has been published in the journal PLOS ONE. Early diagnosis of ASD is vital to make a difference to the lives of young children living with ASD who are typically not diagnosed until the age of four, says Dwight German, PhD, professor of psychiatry at UT Southwestern and senior author of the study. To date, blood-based biomarkers such as neurotransmitters, cytokines, and markers of mitochondrial dysfunction, oxidative stress, and impaired methylation, have been investigated.