Poor taxonomy classification results when compared to querying BOLD db via BOLDigger?

[Anto007]

Hi,

I was wondering why the taxonomy results are poor/inaccurate when using a dada2 formatted coi db using your pipeline versus querying the BOLD coi db via BOLDigger. Below are the taxonomy results that I got after running assignTaxonomy step on dada2 using the coi db file (bold_clustered.assigntaxonomy.fasta) that was generated after running your pipeline (under the default settings on conda)

Kingdom Phylum Class Order Family Genus Species
ASV_1 Animalia Arthropoda Insecta Unclassified Unclassified Unclassified Unclassified
ASV_2 Animalia Unclassified Unclassified Unclassified Unclassified UnclassifiedUnclassified
ASV_3 Animalia Unclassified Unclassified Unclassified Unclassified UnclassifiedUnclassified
ASV_4 Animalia Arthropoda Insecta Unclassified Unclassified Unclassified Unclassified

Below are the results from quering the BOLD coi db via BOLDigger

Phylum Class Order Family Genus Species
ASV_1 Chordata Actinopterygii Gobiiformes Unclassified Unclassified Unclassified
ASV_2 Chordata Actinopterygii Gobiiformes Gobiidae Eviota Unclassified
ASV_3 Chordata Actinopterygii Gobiiformes Gobiidae Eviota Eviota_pardalota
ASV_4 Chordata Actinopterygii Gobiiformes Gobiidae Eviota Eviota_guttata

Here are the example 4 ASV sequences
test_ASVs.fasta.txt
if you would like to test them for yourself.

Thank you in advance for your inputs.