Silva 138.2 reference fastas not at species level for the former Lactobacillus genus

[inespradal]

Hello,

Although the SILVA 138.2 reference fastas are now officially available for DADA2 (see post bellow),

          Silva 138.2 reference fastas are now officially available for DADA2: https://doi.org/10.5281/zenodo.14169026

_Originally posted by @benjjneb in #1983 (comment)

the file silva_nr99_v138.2_toSpecies_trainset.fa does not have the species name for the genus of the former Lactobacillus genus (see below the example of Fructilactobacillus):

is there a reason for it?

[benjjneb]

An important point is that species-level taxonomy is not officially curated by Silva. The species-level reference files we make available combine the genus-species binomial with the official Silva taxonomy down to genus, but since the genus-species binomials are propagated over from the reference accessions, they do not necessarily match the Silva taxonomy.

We deal with this by only including the species name in our digested reference files if it the genus in the genus-species binomial matches the genus from the Silva taxonomy. My guess is that the binomials for these recently renamed *-lactobacillus genera remain "Lactobacillus" in the binomials and thus aren't being propagated to the reference file.

The full code for how the silva release files are digested into the assignTaxonomy reference format is here: https://github.com/benjjneb/dada2/blob/master/R/taxonomy.R#L532