search

tseemann / mlst

:id: Scan contig files against PubMLST typing schemes
GNU General Public License v2.0
198 stars 47 forks source link

Title fight: mlst vs cge #102

Open jdalsdurf opened 3 years ago

jdalsdurf commented 3 years ago

Trying to understand a discrepancy between my much loved mlst vs CGE webtool (https://cge.cbs.dtu.dk/services/).
I have an isolate of S.suis, sequenced with Illumina > denovo assembled contigs generated with unicycler. I am getting 100% match on an allele with CGE (mutS_408) but in mlst I am getting "novel" allele (mutS_317). When I take that novel allele output and run through CGE i get 100% match to CGE allele (mutS_408).

All inputs for mlst are default except for --novel and --csv flags.

Just trying to figure out where i might be generating errors.

love your work, thank you for maintaining mlst.

andersgs commented 3 years ago

I suspect the databases are slightly different. You need to make sure the underlying databases are the same before you can do any comparisons.

tseemann commented 3 years ago

Due to COVID i am 1 year behind on my bacterial software tools.

I have a release ready to go with new databases, but many of the schemes have changed names and that will upset and confused many people.