Machine learning reveals the interplay between muscle fatty acids and the gut microbiota: implications for high-quality pork

Abstract Background High-quality pork is rich in functional fatty acids such as α-linolenic (ALA), γ-linolenic acid (GLA), docosahexaenoic acid (DHA), and eicosapentaenoic acid (EPA), which are thought to prolong lifespan. Multiple studies have confirmed the role of the gut microbiota in muscle fatty acid deposition in pigs, but the inherent complexity of the microbiota remains poorly understood. Results Using a machine learning-based data mining approach, we analyzed a dataset containing meat quality, muscle fatty acids, and gut microbiota from 291 pigs to identify potential predictors of total fatty acid (TFA) and unsaturated fatty acids (UFA) content for high-quality pork with better meat color or marbling score. Microbial α diversity exhibited positive predictive power for muscle polyunsaturated fatty acids (PUFAs), especially for GLA, DHA, and EPA. A core set of two microbial phyla (Actinobacteria and Spirochaetes) and 11 genera extracted from the pig gut microbiome correlated with the level of muscle fatty acids. In addition to the fatty acid degradation pathway, Escherichia was negatively associated with the levels of most fatty acids, including oleic acid, ALA, GLA, and EPA. Positive correlations of GLA, DHA, and EPA depositions in the muscle were shared in two microbial genera (Bradyrhizobium and Lachnoclostridium), which also negatively correlated with the fatty acid degradation pathway. Conclusions Overall, we revealed an intricate network of correlations between muscle fatty acids and the gut microbiota, suggesting that there are multiple routes to produce high-quality healthy pork that is rich in functional fatty acids.