Bioinformatic pipelines have a more considerable impact than sequencing platforms on outputs from shotgun metagenomic analysis of low complexity microbial communities

Shotgun metagenomics has enormous potential for the analysis of low complexity microbial communities, such as those present in foods. Here, we benchmarked the performances of three high-throughput sequencing platforms, the Illumina MiSeq, Illumina NextSeq 500, and Ion Proton, for shotgun metagenomics of food microbiota. Briefly, we sequenced six kefir DNA samples and a mock community DNA sample, which was constructed by mixing genomic DNA from thirteen food-related bacterial species in an equimolar ratio. Subsequently, a variety of bioinformatics tools were used to analyse the data generated, and the effects of sequencing depth on these analyses was tested by randomly subsampling sequences from 100,000 to 7,500,000 reads per sample. Notably, it was apparent that compositional analysis results were consistent between the sequencers at divergent sequencing depths, but clear differences in the outputs from species classifiers were observed, which included differing false-positive rates combined with reference genome size biases. Strain-level analysis results were congruent across the sequencers. Indeed, each sequencer correctly identified the strains present in mock community DNA. However, strain-level analysis accuracy did improve slightly with sequencing depth. For functional analysis, results were again similar for each sequencer at each sequencing depth. In contrasting, metagenome assembly completeness varied hugely at different sequencing depths. Overall, it was apparent that all three sequencing platforms are suitable for shotgun metagenomic analysis of low diversity microbial communities in foods but that care needs to be taken when selecting software for bioinformatic analysis.