25k signatures

[phiweger]

hi @ctb,

I think the "real" GTDB has about 150k genomes, and 25k genomes sounds like a dereplicated set used by the GTDB toolkit, or am I missing something here? Maybe the 25k genomes correspond to this subset by the same group around Donovan Parks?

kind regards,
Adrian

[ctb]

yep, Donovan pointed that out to me this morning! Working on it. The 25k dereplicated seem like a good ~small set to use, tho. More tomorrow.

[phiweger]

I think a small index would be a great idea, bc/ the GTDB "raw" contains about 10% E. coli :)

Just for completeness, we recently dereplicated the GTDB non-manually (compared w/ Parks et al.) see here -- classification accuracy was near-identical.

[ctb]

closing in favor of sourmash-bio/sourmash#2015 and, well, many other things, but sourmash-bio/sourmash#1941 too.