Characterizing the Personalized Microbiota Dynamics for Disease Classification by Individual-Specific Edge-Network Analysis

Conclusion There is growing interest in bolstering resistance to infections or diseases by altering the microbiota (Jia et al., 2008; Holmes, 2016; Waterman et al., 2016; Delzenne and Bindels, 2018). Here, we have presented a computational framework, i.e., iENA, to identify the key OTU features to distinguish normal and disease states, by extracting higher-order statistics and dynamic information from 16S rRNA (ribosomal RNA) gene sequencing data in a one-sample manner. As a proof-of-concept study, we carried out iENA on the temporal development data of twelve subjects (healthy adults) undergoing a challenge with intestinal microbiota by ETEC. Although the sample size is relatively small and the variations among individuals are large, our iENA achieved robust results that may lead to more confirmed conclusions. The analysis outcome from iENA indicates the following: (i) for challenged subjects, the individual symptom-related OTU markers will have stable relation (higher-order information) rather than sensitive OTU abundance; (ii) the OTU markers are significantly related to the disease development and progression (e.g., ETEC infection) which will be able to predict whether an individual would develop symptoms or not with reasonable accuracy. In addition, iENA also showed satisfactory efficiency on another dataset about BV. These consequences all demonstrate the effectiveness of iENA with DNB on an individual’s microbiota dynamics. Excluding the limitations from indivi...
Source: Frontiers in Genetics - Category: Genetics & Stem Cells Source Type: research