Leveraging 16S rRNA Microbiome Sequencing Data to Identify Bacterial Signatures for Irritable Bowel Syndrome

Abstract

Irritable bowel syndrome (IBS) is a chronic gastrointestinal disorder characterized by abdominal pain or discomfort. Previous studies have illustrated that the gut microbiota might play a critical role in IBS, but the conclusions of these studies, based on various methods, were almost impossible to compare, and reproducible microorganism signatures were still in question. To cope with this problem, previously published 16S rRNA gene sequencing data from 439 fecal samples, including 253 IBS samples and 186 control samples, were collected and processed with a uniform bioinformatic pipeline. Although we found no significant differences in community structures between IBS and healthy controls at the amplicon sequence variants (ASV) level, machine learning (ML) approaches enabled us to discriminate IBS from healthy controls at genus level. Linear discriminant analysis effect size (LEfSe) analysis was subsequently used to seek out 97 biomarkers across all studies. Then, we quantified the standardized mean difference (SMDs) for all significant genera identified by LEfSe and ML approaches. Pooled results showed that the SMDs of nine genera had statistical significance, in which the abundance of Lachnoclostridium, Dorea, Erysipelatoclostridium, Prevotella 9, and Clostridium sensu stricto 1 in IBS were higher, while the dominant abundance genera of healthy controls were Ruminococcaceae UCG-005, Holdemanella, Coprococcus 2, and Eubacterium coprostanoligenes group. In summary, based on six published studies, this study identified nine new microbiome biomarkers of IBS, which might be a basis for understanding the key gut microbes associated with IBS, and could be used as potential targets for microbiome-based diagnostics and therapeutics.

Keywords: 16S rRNA; biomarkers; gut microbiome; irritable bowel syndrome; machine learning algorithm.

Figure 1

The top frames represent the three main steps, and the rectangular frame shows the data process. DADA2, divisive amplicon denoising algorithm 2; ASV, amplicon sequence variants; LEfSe, linear discriminant analysis effect size; RF, random forest; IBS, irritable bowel syndrome.

Figure 2

Most of the studies showed microbiome changes, and potential signatures were found at genus-level in each dataset. (A) Left: Area under the ROC curve (AUC) is calculated by the AdaBoost algorithm. Right: Area under the ROC curve (AUC) is calculated by the random forest algorithm. X-axis starts at 0.5. (B) The number of genera biomarkers with p < 0.05 were identified by linear discriminant analysis effect size (LEfSe) analysis.

Figure 3

The AdaBoost and random forest classifiers identified important feature genera in distinguishing IBS from healthy subjects. (A) The important features selected by the AdaBoost classifier. a1. The top 23 genus biomarkers were ranked in descending order of the most relevant features to the model. a2. 10-fold cross-validation score on the Y-axis and the number of features on the X-axis. (B) The important features contributed to random forest corresponding to AdaBoost. b1. 25 important genera sorted in descending order based on mean decrease accuracy (MDA). b2. 10-fold cross-validation error on the Y-axis and the number of features on the X-axis.

Figure 4

Forest plot reporting effect sizes calculated using a meta-analysis of standardized mean differences and a fixed-effects model on nine genera abundances between carcinomas and controls. The SMD had a positive value confirming the higher number of genera in patients with IBS, and a negative value showing more abundance of genera in healthy controls. The length of the error bar depicts the 95% CIs. The fixed model indicates the overall effect sizes SMD value of all studies.