Microbiome signature of posttraumatic stress disorder and resilience in youth.

OBJECTIVE: Identifying biomarkers that can distinguish trauma-exposed youth at risk for developing posttraumatic pathology from resilient individuals is essential for targeted interventions. As trauma can alter the microbiome with lasting effects on the host, our longitudinal, multimeasure, cross-species study aimed to identify the microbial signature of posttraumatic stress disorder (PTSD). METHOD: We followed children exposed to war-related trauma and matched controls from early childhood (Mage = 2.76 years, N = 232) to adolescence (Mage = 16.13 years, N = 84), repeatedly assessing posttraumatic symptomatology and maternal caregiving. In late adolescence, we collected fecal samples from mothers and youth and assessed microbiome composition, diversity, and mother-child microbial synchrony. We then transplanted adolescents' fecal samples into germ-free mice to determine if behavioral changes are observed. RESULTS: Youth with PTSD exhibited a distinct gut microbiome profile and lower diversity compared to resilient individuals, and microbiome diversity mediated the continuity of posttraumatic symptomatology throughout development. Low microbiome diversity correlated with more posttraumatic symptoms in early childhood, more emotional and behavioral problems in adolescence, and poor maternal caregiving. Youth with PTSD demonstrated less mother-child microbial synchrony, suggesting that low microbial concordance between mother and child may indicate susceptibility to posttraumatic illness. Germ-free mice transplanted with microbiomes from individuals with PTSD displayed increased anxious behavior. CONCLUSIONS: Our findings provide evidence that the trauma-associated microbiome profile is at least partially responsible for the anxiety component of the PTSD phenotype and highlight microbial underpinnings of resilience. Further, our results suggest that the microbiome may serve as additional biological memory of early life stress and underscore the potential for microbiome-related diagnosis and treatment following trauma. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

mi-Mic: a novel multi-layer statistical test for microbiota-disease associations.

mi-Mic, a novel approach for microbiome differential abundance analysis, tackles the key challenges of such statistical tests: a large number of tests, sparsity, varying abundance scales, and taxonomic relationships. mi-Mic first converts microbial counts to a cladogram of means. It then applies a priori tests on the upper levels of the cladogram to detect overall relationships. Finally, it performs a Mann-Whitney test on paths that are consistently significant along the cladogram or on the leaves. mi-Mic has much higher true to false positives ratios than existing tests, as measured by a new real-to-shuffle positive score.

Gut microbiome-metabolome interactions predict host condition.

BACKGROUND: The effect of microbes on their human host is often mediated through changes in metabolite concentrations. As such, multiple tools have been proposed to predict metabolite concentrations from microbial taxa frequencies. Such tools typically fail to capture the dependence of the microbiome-metabolite relation on the environment. RESULTS: We propose to treat the microbiome-metabolome relation as the equilibrium of a complex interaction and to relate the host condition to a latent representation of the interaction between the log concentration of the metabolome and the log frequencies of the microbiome. We develop LOCATE (Latent variables Of miCrobiome And meTabolites rElations), a machine learning tool to predict the metabolite concentration from the microbiome composition and produce a latent representation of the interaction. This representation is then used to predict the host condition. LOCATE's accuracy in predicting the metabolome is higher than all current predictors. The metabolite concentration prediction accuracy significantly decreases cross datasets, and cross conditions, especially in 16S data. LOCATE's latent representation predicts the host condition better than either the microbiome or the metabolome. This representation is strongly correlated with host demographics. A significant improvement in accuracy (0.793 vs. 0.724 average accuracy) is obtained even with a small number of metabolite samples ([Formula: see text]). CONCLUSION: These results suggest that a latent representation of the microbiome-metabolome interaction leads to a better association with the host condition than any of the two separated or the simple combination of the two. Video Abstract.

Recipient-independent, high-accuracy FMT-response prediction and optimization in mice and humans.

BACKGROUND: Some microbiota compositions are associated with negative outcomes, including among others, obesity, allergies, and the failure to respond to treatment. Microbiota manipulation or supplementation can restore a community associated with a healthy condition. Such interventions are typically probiotics or fecal microbiota transplantation (FMT). FMT donor selection is currently based on donor phenotype, rather than the anticipated microbiota composition in the recipient and associated health benefits. However, the donor and post-transplant recipient conditions differ drastically. We here propose an algorithm to identify ideal donors and predict the expected outcome of FMT based on donor microbiome alone. We also demonstrate how to optimize FMT for different required outcomes. RESULTS: We show, using multiple microbiome properties, that donor and post-transplant recipient microbiota differ widely and propose a tool to predict the recipient post-transplant condition (engraftment success and clinical outcome), using only the donors' microbiome and, when available, demographics for transplantations from humans to either mice or other humans (with or without antibiotic pre-treatment). We validated the predictor using a de novo FMT experiment highlighting the possibility of choosing transplants that optimize an array of required goals. We then extend the method to characterize a best-planned transplant (bacterial cocktail) by combining the predictor and a generative genetic algorithm (GA). We further show that a limited number of taxa is enough for an FMT to produce a desired microbiome or phenotype. CONCLUSIONS: Off-the-shelf FMT requires recipient-independent optimized FMT selection. Such a transplant can be from an optimal donor or from a cultured set of microbes. We have here shown the feasibility of both types of manipulations in mouse and human recipients. Video Abstract.

Ordering taxa in image convolution networks improves microbiome-based machine learning accuracy.

The human gut microbiome is associated with a large number of disease etiologies. As such, it is a natural candidate for machine-learning-based biomarker development for multiple diseases and conditions. The microbiome is often analyzed using 16S rRNA gene sequencing or shotgun metagenomics. However, several properties of microbial sequence-based studies hinder machine learning (ML), including non-uniform representation, a small number of samples compared with the dimension of each sample, and sparsity of the data, with the majority of taxa present in a small subset of samples. We show here using a graph representation that the cladogram structure is as informative as the taxa frequency. We then suggest a novel method to combine information from different taxa and improve data representation for ML using microbial taxonomy. iMic (image microbiome) translates the microbiome to images through an iterative ordering scheme, and applies convolutional neural networks to the resulting image. We show that iMic has a higher precision in static microbiome gene sequence-based ML than state-of-the-art methods. iMic also facilitates the interpretation of the classifiers through an explainable artificial intelligence (AI) algorithm to iMic to detect taxa relevant to each condition. iMic is then extended to dynamic microbiome samples by translating them to movies.

Sampling bias minimization in disease frequency estimates.

An accurate estimate of the number of infected individuals in any disease is crucial. Current estimates are mainly based on the fraction of positive samples or the total number of positive samples. However, both methods are biased and sensitive to the sampling depth. We here propose an alternative method to use the attributes of each sample to estimate the change in the total number of positive patients in the total population. We present a Bayesian estimator assuming a combination of condition and time-dependent probability of being positive, and mixed implicit-explicit solution for the probability of a person with conditions i at time t of being positive. We use this estimate to predict the total probability of being positive at a given day t. We show that these estimate results are smooth and not sensitive to the properties of the samples. Moreover, these results are a better predictor of future mortality.