Detecting sparse microbial association signals adaptively from longitudinal microbiome data based on generalized estimating equations

multiMiAT: an optimal microbiome-based association test for multicategory phenotypes

Microbes can affect the metabolism and immunity of human body incessantly, and the dysbiosis of human microbiome drives not only the occurrence but also the progression of disease (i.e. multiple statuses of disease). Recently, microbiome-based association tests have been widely developed to detect the association between the microbiome and host phenotype. However, the existing methods have not achieved satisfactory performance in testing the association between the microbiome and ordinal/nominal multicategory phenotypes (e.g. disease severity and tumor subtype). In this paper, we propose an optimal microbiome-based association test for multicategory phenotypes, namely, multiMiAT. Specifically, under the multinomial logit model framework, we first introduce a microbiome regression-based kernel association test for multicategory phenotypes (multiMiRKAT). As a data-driven optimal test, multiMiAT then integrates multiMiRKAT, score test and MiRKAT-MC to maintain excellent performance in diverse association patterns. Massive simulation experiments prove the success of our method. Furthermore, multiMiAT is also applied to real microbiome data experiments to detect the association between the gut microbiome and clinical statuses of colorectal cancer as well as for diverse statuses of Clostridium difficile infections.

CDEMI: characterizing differences in microbial composition and function in microbiome data

Microbial communities influence host phenotypes through microbiota-derived metabolites and interactions between exogenous active substances (EASs) and the microbiota. Owing to the high dynamics of microbial community composition and difficulty in microbial functional analysis, the identification of mechanistic links between individual microbes and host phenotypes is complex. Thus, it is important to characterize variations in microbial composition across various conditions (for example, topographical locations, times, physiological and pathological conditions, and populations of different ethnicities) in microbiome studies. However, no web server is currently available to facilitate such characterization. Moreover, accurately annotating the functions of microbes and investigating the possible factors that shape microbial function are critical for discovering links between microbes and host phenotypes. Herein, an online tool, CDEMI, is introduced to discover microbial composition variations across different conditions, and five types of microbe libraries are provided to comprehensively characterize the functionality of microbes from different perspectives. These collective microbe libraries include (1) microbial functional pathways, (2) disease associations with microbes, (3) EASs associations with microbes, (4) bioactive microbial metabolites, and (5) human body habitats. In summary, CDEMI is unique in that it can reveal microbial patterns in distributions/compositions across different conditions and facilitate biological interpretations based on diverse microbe libraries. CDEMI is accessible at http://rdblab.cn/cdemi/.

Characterization of longitudinal nasopharyngeal microbiome patterns in maternally HIV-exposed Zambian infants

Background: Previous studies of infants born to HIV-positive mothers have linked HIV exposure to poor outcomes from gastrointestinal and respiratory illnesses, and to overall increased mortality rates. The mechanism behind this is unknown, but it is possible that differences in the nasopharyngeal (NP) microbiome between HIV-unexposed and HIV-exposed infants could play a role in perpetuating some outcomes. Methods: We conducted a longitudinal analysis of 170 NP swabs of healthy HIV-exposed, uninfected (HEU; n=10) infants and their HIV(+) mothers and HIV-unexposed, uninfected (HUU; n=10) infants and their HIV(-) mothers. These swabs were identified from a sample library collected in Lusaka, Zambia between 2015 and 2016. Using 16S rRNA gene sequencing, we characterized the maturation of the microbiome over the first 14 weeks of life to determine what quantifiable differences exist between HEU and HUU infants, and what patterns are reflected in the mothers' NP microbiomes. Results: In both HEU and HUU infants, Staphylococcus and Corynebacterium began as primary colonizers of the NP microbiome but were in time replaced by Dolosigranulum, Streptococcus, Moraxella and Haemophilus. When studying differences between infants, the microbe Staphylococcus haemolyticus indicated a distinctive high association with HIV exposure at birth, even when accounting for the interaction between HIV exposure status and time of sampling. When comparing infants to their mothers with paired analyses, HEU infants’ NP microbiome composition was only slightly different from their HIV(+) mothers at birth or 14 weeks, including in their carriage of S. pneumoniae, H. influenzae, and S. haemolyticus. Conclusions: Our analyses indicate that the HEU infants in our study exhibit subtle differences in the NP microbial composition throughout the sampling interval. Given our results and the sampling limitations of our study, we believe that further research must be conducted in order to confidently understand the relationship between HIV exposure and infants’ NP microbiomes.