Potential identification of vitamin B6 responsiveness in autism spectrum disorder utilizing phenotype variables and machine learning methods

Abstract

We investigated whether machine learning methods could potentially identify a subgroup of persons with autism spectrum disorder (ASD) who show vitamin B6 responsiveness by selected phenotype variables. We analyzed the existing data from our intervention study with 17 persons. First, we focused on signs and biomarkers that have been identified as candidates for vitamin B6 responsiveness indicators. Second, we conducted hypothesis testing among these selected variables and their combinations. Finally, we further investigated the results by conducting cluster analyses with two different algorithms, affinity propagation and k-medoids. Statistically significant variables for vitamin B6 responsiveness, including combination of hypersensitivity to sound and clumsiness, and plasma glutamine level, were included. As an a priori variable, the Pervasive Developmental Disorders Autism Society Japan Rating Scale (PARS) scores was also included. The affinity propagation analysis showed good classification of three potential vitamin B6-responsive persons with ASD. The k-medoids analysis also showed good classification. To our knowledge, this is the first study to attempt to identify subgroup of persons with ASD who show specific treatment responsiveness using selected phenotype variables. We applied machine learning methods to further investigate these variables’ ability to identify this subgroup of ASD, even when only a small sample size was available.

Prospective associations of the infant gut microbiome and microbial function with social behaviors related to autism at age 3 years

“…Our findings regarding the superpathway of pyridoxal 5′-phosphate (highly available vitamin B6) biosynthesis and salvage are relevant to clinical trials examining the effectiveness of high-dose vitamin B6 and magnesium supplementation in reducing autistic symptoms. While most of these studies have been small, and with conflicting results 55 – 59 , a more recent study suggests that a certain phenotypic subset of autistic subjects respond to B6 and magnesium supplementation 60 . Further research is needed to clarify whether/what gut microbial functions can identify responders so that treatment can be more successfully targeted.…”

Clustering by phenotype and genome-wide association study in autism

“…The SSC data were publicly released in October 2007 and are directly available from the SFARI. From the SSC data set, we used data from 614 affected white male probands who had no missing information regarding Autism Diagnostic Interview-Revised (ADI-R) scores 15 and vitamin treatment 16 , 17 and 391 unaffected brothers for whom genotype data, generated by the Illumina Human Omni2.5 (Omni2.5) array, were available for subsequent clustering and genetic analyses. We excluded participants whose ancestries were estimated to be different from the other participants using principal component analyses (PCAs) performed by EIGENSOFT version 7.2.1 18 for the genotype data.…”

“…We conducted cluster analyses using phenotypic variables of ADI-R scores and history of vitamin treatment 16 , 17 . We chose these variables because the ADI-R is one of the most reliable estimates of ASD and has the ability to evaluate substructure domains of ASD 15 .…”

“…Among the treatments, we selected the variable of history of vitamin treatment because we recently found that a cluster of persons with ASD is associated with potential responsiveness to vitamin B6 treatment 16 , 17 . The history of treatment is not always compatible with responsiveness, but we considered that continuous treatment indicates responsiveness to some degree.…”

Artificial intelligence and machine learning‐aided drug discovery in central nervous system diseases: State‐of‐the‐arts and future directions

“…Also, there are few ML applications in drug discovery for ASD. For example, ML‐based cluster analysis (i.e., affinity propagation and k‐medoids) of clinical data (i.e., signs and biomarkers) exhibited a good performance in drug response prediction of ASD patients 372 . Moreover, Bayesian ML models trained on HTS data revealed the potential repurposing of nicardipine or other dihydropyridine calcium channel antagonists for the treatment of Pitt Hopkins Syndrome, a rare genetic disorder that exhibits features of autistic spectrum disorders 373 .…”