Classification and adaptive behavior prediction of children with autism spectrum disorder based upon multivariate data analysis of markers of oxidative stress and DNA methylation
PLoS Comput Biol. 2017 Mar 16;13(3):e1005385. doi: 10.1371/journal.pcbi.1005385. eCollection 2017 Mar.
Classification and adaptive behavior prediction of children with autism spectrum disorder based upon multivariate data analysis of markers of oxidative stress and DNA methylation
Daniel P. Howsmon, Uwe Kruger, Stepan Melnyk, S. Jill James, Juergen Hahn
Published: March 16, 2017, https://doi.org/10.1371/journal.pcbi.1005385
Daniel P. Howsmon
Affiliations Department of Chemical and Biological Engineering, Rensselaer Polytechnic Institute, Troy, New York, United States of America, Center for Biotechnology and Interdisciplinary Studies, Rensselaer Polytechnic Institute, Troy, New York, United States of America
ORCID logo http://orcid.org/0000-0002-7177-1342
Uwe Kruger
Affiliation Department of Biomedical Engineering, Rensselaer Polytechnic Institute, Troy, New York, United States of America
ORCID logo http://orcid.org/0000-0001-5664-9499
Stepan Melnyk
Affiliation Department of Pediatrics, University of Arkansas for Medical Sciences, Little Rock, Arkansas, United States of America
S. Jill James
Affiliation Department of Pediatrics, University of Arkansas for Medical Sciences, Little Rock, Arkansas, United States of America
Juergen Hahn
E-mail: hahnj@rpi.edu
Affiliations Department of Chemical and Biological Engineering, Rensselaer Polytechnic Institute, Troy, New York, United States of America, Center for Biotechnology and Interdisciplinary Studies, Rensselaer Polytechnic Institute, Troy, New York, United States of America, Department of Biomedical Engineering, Rensselaer Polytechnic Institute, Troy, New York, United States of America
Abstract
The number of diagnosed cases of Autism Spectrum Disorders (ASD) has increased dramatically over the last four decades; however, there is still considerable debate regarding the underlying pathophysiology of ASD. This lack of biological knowledge restricts diagnoses to be made based on behavioral observations and psychometric tools. However, physiological measurements should support these behavioral diagnoses in the future in order to enable earlier and more accurate diagnoses. Stepping towards this goal of incorporating biochemical data into ASD diagnosis, this paper analyzes measurements of metabolite concentrations of the folate-dependent one-carbon metabolism and transulfuration pathways taken from blood samples of 83 participants with ASD and 76 age-matched neurotypical peers. Fisher Discriminant Analysis enables multivariate classification of the participants as on the spectrum or neurotypical which results in 96.1% of all neurotypical participants being correctly identified as such while still correctly identifying 97.6% of the ASD cohort. Furthermore, kernel partial least squares is used to predict adaptive behavior, as measured by the Vineland Adaptive Behavior Composite score, where measurement of five metabolites of the pathways was sufficient to predict the Vineland score with an R2 of 0.45 after cross-validation. This level of accuracy for classification as well as severity prediction far exceeds any other approach in this field and is a strong indicator that the metabolites under consideration are strongly correlated with an ASD diagnosis but also that the statistical analysis used here offers tremendous potential for extracting important information from complex biochemical data sets.
