Machine Learning-Aided Meta-Analysis Reveals Changes in Penaeus vannamei Gut Bacterial Communities Upon Dietary...

Gut bacterial communities play a key role in shrimp health; thus, their modulation has been a target of dietary supplements which also function in enhancing disease and stress resistance of shrimp. However, this also raised the question of whether immunostimulants yield distinct changes in the gut bacterial composition or whether there are consistent features across all treatments. Here, we performed a machine learning-aided meta-analysis of 16S rRNA gut bacterial community studies of immunostimulants for Penaeus vannamei. Sequence reads from the selected studies were obtained and processed through bioinformatics tools. While beta diversity analysis suggests similarities between the normal, infected and stimulated shrimp, alpha diversity indices showed higher species richness in the gut bacterial communities of shrimp fed with immunostimulants. Specific beneficial taxa were enriched upon immunostimulation, while potentially pathogenic taxa decreased in abundance. Random forest modelling also identified key predictor taxa which may be used to classify gut bacterial communities based on immune status, type of immunostimulant and the specific immunostimulant. Despite some shared patterns in differential abundance-having decreased relative abundances of Photobacterium and other members of Gammaproteobacteria-the influence of immunostimulation on gut bacterial community composition was type- and treatment-specific, as evident in the distinct abundance profiles of the predictor taxa. Functional prediction analysis also showed distinct pathways enriched in immunostimulated shrimp, as influenced by the type of the immunostimulant. This study highlighted the specific impacts of dietary supplementation-induced immunostimulation on gut bacterial communities and identified key features in immunostimulated shrimp which provide a novel perspective on the interplay between gut bacterial community and immunity.