A meta-analysis study of the robustness and universality of gut microbiota-shrimp diseases relationship.

ABSTRACT

Intensive case study has established dysbiosis in the gut microbiota-shrimp disease relationship; however, variability in experimental design and the diversity of diseases arise the question of whether some gut indicators are robust and universal in response to shrimp health status, irrespective of causal agents. Through an unbiased subject-level meta-analysis framework, we re-analysed 10 studies, including 261 samples, four lifestages and six different diseases (the causal agents are virus, bacterial, eukaryotic pathogens, or unknown). Results showed that shrimp diseases reproducibly altered the structure of gut bacterial community, but not diversity. After ruling out the lifestage- and disease specific- discriminatory taxa (different diseases dependent indicators), we identify 18 common disease-discriminatory taxa (indicative of health status, irrespective of causal agents) that accurately diagnosed (90.0% accuracy) shrimp health status, regardless of different diseases. These optimizations substantially improved the performance (62.6% vs. 90.0%) diagnosing model. The robustness and universality of model were validated for effectiveness via leave-one-dataset-out validation and independent cohorts. Interspecies interaction and stability of the gut microbiotas were consistently compromised in diseased shrimp compared with corresponding healthy cohorts, while stochasticity and beta-dispersion exhibited the opposite trend. Collectively, our findings exemplify the utility of microbiome meta-analyses in identifying robust and reproducible features for quantitatively diagnosing disease incidence, and the downstream consequences for shrimp pathogenesis from an ecological prospective.