| Popis: |
BackgroundUnderstanding the composition and function of the microbiome is essential to understanding vertebrates. The taxonomic composition of a microbiome is often identified using amplicon sequencing of the 16S rRNA gene, but as a single marker, it cannot identify functions (genes). Metagenome and metatranscriptome sequencing can determine microbiome function but can be cost prohibitive. Therefore, computational methods have been developed to generate simulated metagenomes derived from 16S rRNA sequences and full-length genomes. Simulated metagenomes can be an effective alternative to empirical sequencing, but accuracy depends on the genomic database used and whether the database contains organisms closely related to the 16S sequences. The effectiveness of simulated metagenomes in non-model systems is poorly known. We sought to determine the accuracy of simulated metagenomes in a non-model organism, the Canada goose (Branta canadensis), by comparing metagenomes and metatranscriptomes to simulated metagenomes derived from 16S amplicon sequencing of the same samples. ResultsThe Canada goose fecal microbiome is rich in microbial taxa and functions, including those involved in fermentation. There were significant differences between the metagenomes, metatranscriptomes, and simulated metagenomes when comparing enzymes (ECs), KEGG orthologies (KOs), and metabolic pathways. The simulated metagenomes, when compared to the metagenomes, accurately identified the majority of the total ECs, KOs and pathways. The simulated metagenomes accurately identified the majority of the short-chain fatty acid metabolic pathways crucial to organisms with a nutrient poor diet. When narrowed in scope to specific groups of genes, the simulated metagenomes overestimated the number antimicrobial resistance genes and underestimated the number of genes relating to digestion in folivores.ConclusionsOur data show simulated metagenomes may be a useful tool when studying the functional potential of a non-model organism’s microbiome, depending on the question being asked. Simulated metagenomes were selectively accurate when identifying certain characteristics of the microbiome, like metabolic pathways, but not when compared to the entire sequenced metagenome. While simulated metagenomes are an important tool for inferring function, the lack of whole microbial genomes from understudied systems provides an impetus for continued study and sequencing of whole genomes and metagenomes from non-model systems. |
