Hey Daniel
Thank you for adding the GTDB mapping file to the list of mapping files!
I was comparing the manual GTDB blast results of one contig with MEGAN LCA method mapping to the GTDB database. I have a contig in which all the protein sequences match to the same microbe at the species level, giving me an opportunity to examine the sensitivity of taxonomic annotation at a lower taxonomic level.
Blastp analysis against GTDB database shows that the “consensus” taxonomy of the contig is-
“d__Bacteria_p__Fibrobacterota_c__Chitinivibrionia_o__Chitinivibrionales_f__Chitinispirillaceae_g__GUT77”. Consensus means that all 8 proteins in the contig have the same taxonomic annotation.
While in MEGAN LCA analysis, whereby I performed blastx of the contig to nr database and generated GTDB taxonomy by “–read2class GTDB” flag, taxonomy is-
“GTDB;Bacteria;Fibrobacterota;Chitinivibrionia;Chitinivibrionales;Chitinispirillaceae;Chitinispirillum;Chitinispirillum alkaliphilum;Chitinispirillum alkaliphilum ACht6-1”
The outputs match at the family level so they are not completely different. But as there is no potential contamination in the contig, taxonomic classification at the genus level should match, right?
I am using the command line version of MEGAN version 6.21.1.
Looking forward to your reply!