Metabolome-informed microbiome analysis refines metadata classifications and reveals unexpected medication transfer in captive cheetahs
Structured AbstractEven high-quality collection and reporting of study metadata in microbiome studies can lead to various forms of inadvertently missing or mischaracterized information that can alter the interpretation or outcome of the studies, especially with non-model organisms. Metabolomic profiling of fecal microbiome samples can provide empirical insight into unanticipated confounding factors that are not possible to obtain even from detailed care records. We illustrate this point using data from cheetahs from the San Diego Zoo Safari Park. The metabolomic characterization indicated that one cheetah had to be moved from the non-antibiotic-exposed to the antibiotic-exposed group. The detection of the antibiotic in this second cheetah was likely due to grooming interactions with the cheetah that was administered antibiotics. Similarly, because transit time for stool is variable, early fecal samples within the first few days of antibiotic prescription do not all contain detectable antibiotics. Therefore, the microbiome is not affected by the antibiotics at those time points. These insights significantly altered the way the samples were grouped for analysis (antibiotic vs no antibiotic), and the subsequent understanding of the effect of the antibiotics on the cheetah microbiome. Metabolomics also revealed information about numerous other medications, and provided unexpected dietary insights that in turn improved our understanding of the molecular patterns on the impact on the community microbial structure. These results suggest that untargeted metabolomics data provide empirical evidence to correct records of non-model organisms in captivity, although we also expect these methods will be appropriate for experimental conditions typical in human studies.\n\nImportanceMetabolome-informed analyses can enhance omics studies by enabling the correct partitioning of samples by identifying hidden confounders inadvertently misrepresented or omitted from carefully curated metadata. We demonstrate the utility of metabolomics in a study characterizing the microbiome associated with liver disease in cheetahs. Metabolome-informed reinterpretation of metagenome and metabolome profiles factored in an unexpected transfer of antibiotics preventing misinterpretation of the data. Our work suggests that untargeted metabolomics can be used to verify, augment, and correct sample metadata to support improved grouping of sample data for microbiome analyses, here for non-model organisms in captivity. However, the techniques also suggest a path forward for correcting clinical information in human studies to enable higher-precision analyses.