Seems like a super useful tool! 🖥️🧬🦠 "A fast comparative genome browser for diverse bacteria and archaea" preprint: https://www.biorxiv.org/content/10.1101/2023.08.23.554478v1https://fast.genomics.lbl.gov/cgi/search.cgi
note that the default thresholds for the second pass were selected somewhat arbitrary. You could play with them if it looks like pathways are missing genes for certain steps, or just to see whether things get better/worse that way.
Genetic manipulation of Patescibacteria provides mechanistic insights into microbial dark matter and the epibiotic lifestyle www.cell.com/cell/fulltex... #jcampubs
Discovery and exploitation of natural competence in Patescibacteria enables visualization of the bacterial epibiont lifecycle and dissection of its genetic underpinnings.
And of course, the addiitonal annotations are a nice feature of anvi'o, alhough it should be said that kofamscan also includes a "--threshold-scale" parameter allowing the user to modfiy the threshold there as well, although i think the anvio heuristic is more elegant. 8/8
This was in the (excellent!) anvi'o documentation all along, but I did not realise it, despite having used the anvi-run-kegg-kofams program somewhat frequently. A cautionary tale for me to check docs better as tools get updated to newer versions 7/8
Detailed documentation of the heuristic is here: https://anvio.org/help/main/programs/anvi-run-kegg-kofams/https://github.com/merenlab/anvio/blob/master/anvio/kegg.py#L1906-L2031 6/8
It turns out that the anvi'o team (in this case Iva Veseli) has implemented a heuristic to find additional hits in case the default cutoffs of the hmms included in kofamscan database are too stringent. There's a second pass over the genes that fell just below the cutoff values 5/8
When I tried this on a Nirospira lineage 1 genome (related to model organism Nitrospira defluvii), the difference was much less pronounced (kofamscan 1838 vs anvi'o 2061), but still more than 10% extra genes recieved an annotation from anvi'o 4/8
The person asking the question had compared the KOfam annotation of their genome using anvi'o with kofamscan & microbeannotator (same DB version) and found that anvi'o annotated roughly 1900 genes in the genome, whereas the other two annotated ~700. Quite the difference! 3/8
I quite like anvi'o, by Meren and colleagues, for microbial (meta)genome analyses and follow the anvi'o discord server where the team addresses user questions (invite link on https://anvio.org/). Recently a question came up asking details about how anvi'o handles kegg annotations. 2/8