Large 16S + Metaphlan discrepancies

Hello. I have some soil samples that I have performed 16S rRNA amplicon sequencing and metagenomics for. However, I appear to be getting quite large differences in the taxa that are present / identified. For the metagenomics, it was sequences reasonably deeply (roughly 4.2 gigabases of cleaned Kneaddata output data; ~10 gigabytes file size; related question is #3105).

For the 16S data, I have a rough species richness of 130. The Metaphlan3 using mpa_v30_CHOCOPhlAn_201901 (default parameters) is only showing about 15 species. Metaphlan3 is also running very quickly (~30 minutes), is that unusual? I saw a benchmark study and the median time was around 3 hours… I reran Metaphlan using --stat_q 0.1 and there’s now 40 species present. One complete distinction is the near lack of Nitro- bacteria in the metagenomics data (one species was present in the stat_q = 0.1) compared to the 16S data. Is this a typical output from a soil environment or am I doing something wrong?


1 Like