Incertae_sedis VS NA Taxonomic Assignment

[mya-darsan]

Hello!

I used UNITE to assign taxonomy. When looking at the assignments some received:

• OTU_6 k__Fungi NA NA NA NA NA NA
• OTU_4 kFungi pFungi_phy_Incertae_sedis cFungi_cls_Incertae_sedis o__Fungi_ord_Incertae_sedis fFungi_fam_Incertae_sedis g__Fungi_gen_Incertae_sedis NA

I am wondering what the difference is between an OTU being assigned NA and Incertae_sedis?

Is NA just completely not found in the database (therefore not a fungus even though it get kingdom fungus?) and Incertae_sedis is when it is a fungus but the relationship is unknown?

[benjjneb]

NA is assigned when the taxaonomic classification method finds different assignments at a given taxonomic level from subsets of the sequence than it did from the full sequence. The specifics of how this is done are described in the original paper on the naive Bayesian classifier: https://doi.org/10.1128/AEM.00062-07

An Incertae_sedis assignment means that the taxonomic classification method found a reference sequences with Incertae_sedis at that taxonomic level from the full length sequences, and from most of the subsets of the sequences. So, in some sense this "classification" is confident. However, Incertae_sedis means that the taxonomic placement at that level is uncertain. Thus, I would generally interpret this as the same as an NA -- we don't know the classification.

In shorter version: NA comes from uncertainty at the level of comparing the query sequence to the reference database, while Incertae_sedis comes from uncertainty of taxonomic assignments for the reference database entries themselves.