The Potential of Salivary Lipid-Based Cannabis-Responsive Biomarkers to Evaluate Medical Cannabis Treatment in Children with Autism Spectrum Disorder

A machine learning approach for understanding the metabolomics response of children with autism spectrum disorder to medical cannabis treatment

“…Fifteen children (average age 9.4 years) participated in the ASD group and 10 children with a similar age distribution (average age 9.3 years) participated in the TD untreated control group as previously described in detail 14 , 15 . Within the ASD group 11 children exhibited severe range, 2 children exhibited moderate range and 2 children exhibited a mild range of social impairment associated with ASD as reported by parent ratings (SRS-2), as described in detail by Siani-Rose et al 14 (Suppl 2).…”

“…We applied ML applications to source data for each child containing the absolute values of 645 metabolites detected in the saliva collected from study participants by dual scan capillary electrophoresis time-of-flight-mass spectrometry (CE–TOF–MS) and rapid resolution liquid chromatography-time-of-flight-mass spectrometry (RRLC–TOF–MS) 14 , 15 . The samples were designated in children with ASD before MC treatment as PRE and approximately 90 min after MC treatment as PEAK, and a single sample from children in the untreated TD control group as (TD).…”

“…We have recently reported on a new pharmacometabolomics approach in which the levels of metabolites, specifically cannabis-responsive biomarkers, in children with ASD being treated with MC shifted toward the physiologic levels detected in untreated typically developing (TD) children 14 , 15 . In these investigations, we identified a possible link between cannabis-responsive biomarkers and mitochondrial dysfunction, alterations in neurotransmitters, abnormal neuronal development, neuroinflammation, bioenergy and oxidative stress.…”

“…In our previous studies 14 , 15 we used hard-coded algorithms to identify and rank biomarkers that solely shift toward physiological levels. Here we explore the potential of the cannabis-responsive biomarker database, which contains a large number of metabolites detected in a limited number of children with ASD, regardless of the physiological outcomes, and TD controls, for ML applications primarily yielding new biomarker candidates overlooked in previous studies.…”

The Importance of Data Sources for Machine Learning Applications in Autism: A Mini Review

“…A recent study by Quillet et al 25 identified biomarkers distinguishing ASD and typically developing (TD) groups and linked the cannabinoids tetrahydrocannabinol (THC), cannabidiol (CBD) and cannabigerol (CBG) with metabolite levels in children with ASD. It was the first to use ML algorithms on a pharmacometabolomics dataset of previously identified cannabis-responsive biomarkers and other metabolites in children with ASD that shift toward physiological levels determined in typically developing children (TD) after successful medical cannabis treatment 23,26 .…”