Analysing microbial communities with next-generation sequencing (NGS) is unveiling diagnostic and therapeutic biomarkers for various diseases. The NGS approach 16S rRNA amplicon sequencing (16S) offers an economically viable option to characterise microbial community for these types of studies. However, 16S is limited by detail and accuracy for microbial taxonomic hierarchy and gene predictions. In contrast, metagenomic shotgun sequencing overcomes these limitations but at a far greater cost. Previous studies have compared 16S and shotgun techniques but were restricted by small sample sizes (n=1), were conducted without biological context and were limited or without statistical analysis.
We aim to investigate the gut microbial response to a proposed Anti-Interleukin (IL)-23 treatment for Spondyloarthropathy (SpA) diseases. This study applied four different treatment regimens to 32 SKG mice and faecal samples were collected at three different time points during regime stages. These samples were analysed with 16S and shotgun sequencing techniques to examine how each approach represents gut microbial communities of the different treatment groups.
In this study we compare each treatment groups’ microbial community as well as gene annotations using alpha-diversity, beta-diversity along with a statistical power assessment. More specifically we applied the spares Partial Least Squares Discriminant Analysis (sPLS-DA), which characterises microbial and gene signatures that discriminates treatment groups of different time points and generates graphical outputs. We also propose recommendation about sample size and statistical power for the two sequencing methods. Comparisons were conducted with 16S and shotgun sequencing approaches. Taxonomy classifications were conducted with QIIME2 and GraftM and gene annotations were generated with PICRUSt and MG-RAST.
Both sequencing approaches identified similar family ranking classification for high abundance OTUs. Although, 16S proved to have reduced detail and accuracy in taxonomy hierarchy and gene annotations. Overall, shotgun had greater taxonomy and gene resolution. The shotgun approach also performed better at identifying microbial and gene signatures that discriminated treatment groups and maintained a higher statistical power for both taxonomy and gene annotations.