This question goes for MetaPhlAn version 4.0.6 (1 Mar 2023) but could be related to any MetaPhlAn versions.
When obtaining the results the profiled .tsv obtains the columns “clade”, “NCBI_tax_id”, “relative_abundance” and “additional species”. Regarding the “NCBI_tax_id” column, it seems that the MetaPhlAn mpa_vJun23_CHOCOPhlAnSGB_202403, (the one I have installed) is related to the NCBI database.
Is there any way to use a SILVA related database? or to convert silva-138-99-nb-classifier.qza or any other *.qza database into a MetaPhlAn compatible format?
Hi @PARCB,
Indeed the taxonomies reported come from NCBI. The reason is that the database is built from genomes from NCBI and other genomes assembled from metagenomic samples, which are clustered into species-level genome bins (SGBs) at 95% Average Nucleotide Identity, from which we extract SGB-specific marker genes (see Extending and improving metagenomic taxonomic profiling with uncharacterized species using MetaPhlAn 4 | Nature Biotechnology). Due to the way the database is built and the code implemented there is no easy way to use an external rrna based database.