Utilizing stability criteria in choosing feature selection methods yields reproducible results in microbiome data

Feature selection is indispensable in microbiome data analysis, but it can be particularly challenging as microbiome data sets are high dimensional, underdetermined, sparse and compositional. Great efforts have recently been made on developing new methods for feature selection that handle the above data characteristics, but almost all methods were evaluated based on performance of model predictions. However, little attention has been paid to address a fundamental question: how appropriate are those evaluation criteria? Most feature selection methods often control the model fit, but the ability to identify meaningful subsets of features cannot be evaluated simply based on the prediction accuracy. If tiny changes to the data would lead to large changes in the chosen feature subset, then many selected features are likely to be a data artifact rather than real biological signal. This crucial need of identifying relevant and reproducible features motivated the reproducibility evaluation criterion such as Stability, which quantifies how robust a method is to perturbations in the data. In our paper, we compare the performance of popular model prediction metrics (MSE or AUC) with proposed reproducibility criterion Stability in evaluating four widely used feature selection methods in both simulations and experimental microbiome applications with continuous or binary outcomes. We conclude that Stability is a preferred feature selection criterion over model prediction metrics because it better quantifies the reproducibility of the feature selection method.

Suppression of the gut microbiome ameliorates age-related arterial dysfunction and oxidative stress in mice

KEY POINTS

Age-related arterial dysfunction, characterized by oxidative stress- and inflammation-mediated endothelial dysfunction and arterial stiffening, is the primary risk factor for cardiovascular diseases. To investigate whether age-related changes in the gut microbiome may mediate arterial dysfunction, we suppressed gut microbiota in young and old mice with a cocktail of broad-spectrum, poorly-absorbed antibiotics in drinking water for 3-4 weeks. In old mice, antibiotic treatment reversed endothelial dysfunction and arterial stiffening and attenuated vascular oxidative stress and inflammation. To provide insight into age-related changes in gut microbiota that may underlie these observations, we show that ageing altered the abundance of microbial taxa associated with gut dysbiosis and increased plasma levels of the adverse gut-derived metabolite trimethylamine N-oxide. The results of the present study provide the first proof-of-concept evidence that the gut microbiome is an important mediator of age-related arterial dysfunction and therefore may be a promising therapeutic target for preserving arterial function with ageing, thereby reducing the risk of cardiovascular diseases.

ABSTRACT

Oxidative stress-mediated arterial dysfunction (e.g. endothelial dysfunction and large elastic artery stiffening) is the primary mechanism driving age-related cardiovascular diseases. Accumulating evidence suggests the gut microbiome modulates host physiology because dysregulation ('gut dysbiosis') has systemic consequences, including promotion of oxidative stress. The present study aimed to determine whether the gut microbiome modulates arterial function with ageing. We measured arterial function in young and older mice after 3-4 weeks of treatment with broad-spectrum, poorly-absorbed antibiotics to suppress the gut microbiome. To identify potential mechanistic links between the gut microbiome and age-related arterial dysfunction, we sequenced microbiota from young and older mice and measured plasma levels of the adverse gut-derived metabolite trimethylamine N-oxide (TMAO). In old mice, antibiotics reversed endothelial dysfunction [area-under-the-curve carotid artery dilatation to acetylcholine in young: 345 ± 16 AU vs. old control (OC): 220 ± 34 AU, P < 0.01; vs. old antibiotic-treated (OA): 334 ± 15 AU; P < 0.01 vs. OC] and arterial stiffening (aortic pulse wave velocity in young: 3.62 ± 0.15 m  s-1  vs. OC: 4.43 ± 0.38 m  s-1 ; vs. OA: 3.52 ± 0.35 m  s-1 ; P = 0.03). These improvements were accompanied by lower oxidative stress and greater antioxidant enzyme expression. Ageing altered the abundance of gut microbial taxa associated with gut dysbiosis. Lastly, plasma TMAO was higher with ageing (young: 2.6 ± 0.4 μmol  L-1   vs. OC: 7.2 ± 2.0 μmol  L-1 ; P < 0.0001) and suppressed by antibiotic treatment (OA: 1.2 ± 0.2 μmol  L-1 ; P < 0.0001 vs. OC). The results of the present study provide the first evidence for the gut microbiome being an important mediator of age-related arterial dysfunction and oxidative stress and suggest that therapeutic strategies targeting gut microbiome health may hold promise for preserving arterial function and reducing cardiovascular risk with ageing in humans.