Leveraging Existing 16SrRNA Microbial Data to Define a Composite Biomarker for Autism Spectrum Disorder.

Cumulative studies have utilized high-throughput sequencing of the 16SrRNA gene to characterize the composition and structure of the microbiota in autism spectrum disorder (ASD). However, they do not always obtain consistent results; thus, conducting cross-study comparisons is necessary. This study sought to analyze the alteration of fecal microbiota and the diagnostic capabilities of gut microbiota biomarkers in individuals with ASD using the existing 16SrRNA microbial data and explore heterogeneity among studies. The raw sequence and metadata from 10 studies, including 1,019 samples, were reanalyzed. Results showed no significant difference in alpha diversity of fecal microbiota between ASD and the control group. However, a significant difference in the composition structure of fecal microbiota was observed. Given the large differences in sample selection and technical differences, the separation of fecal microbiota between ASD and controls was not observed. Subgroup analysis was performed on the basis of different country of origin, hypervariable regions, and sequencing platforms, and the dominant genera in ASD and healthy control groups were determined by linear discriminant analysis (LDA) of the effect size (LEfSe) algorithm and Wilcoxon rank-sum test. Machine learning analyses were carried out to determine the diagnostic capabilities of potential microbial biomarkers. A total of 12 genera were identified to distinguish ASD from control, and the AUC of the training set and verification set was 0.757 and 0.761, respectively. Despite cohort heterogeneity, gut microbial dysbiosis of ASD has been proven to be a widespread phenomenon. Therefore, fecal microbial markers are of great significance in diagnosing ASD diseases and possible candidates for further mechanistic study of the role of intestinal microbiota in ASD.