Methods

Methods

How a call is made

The sizes an assay's profiles use are first merged into canonical bands: sizes within --tolerance of each other become one band, because no gel separates them. Each sample is then reduced to the presence or absence of every canonical band, and that pattern is compared to the profiles in brucetyper/assays.py:

• Bruce-ladder and Suis-ladder match exactly — every canonical band must agree in presence and absence.
• AMOS and the RB51 junction PCR match on a subset — the bands of the profile must be present, extra bands are tolerated, and the first profile listed that fits wins.

A sample matching no profile is called N/A. When an assay genuinely cannot separate two genotypes, both names are reported joined by /.

Repeated loci are collapsed before calling: a locus present several times in a genome yields several products of the same size, which a gel shows as one band.

All primers and expected band sizes live in brucetyper/assays.py and nowhere else. Edit that file to change how genotypes are called. Every primer has been checked against its publication and every band size against the product it gives in 87 real genomes: see Validation.

The rough flag

Bruce-ladder rows carry a rough column, read from one band and one band only: the 1682 bp product of BMEI0998_BMEI0997, which spans wboA and wboB, the two glycosyltransferases of the O-antigen locus. no means the band is intact, yes means it is shifted or missing, and the column is empty for Suis-ladder, AMOS and the RB51 junction PCR, which do not carry that band. The flag is reported beside the call and never changes it.

It is a reading of one band, not a phenotype:

• It misses B. canis. B. canis is naturally rough but keeps an intact 1682 bp band, so all ten canis genomes tested are flagged no. Its roughness does not come from losing wboA.
• An incomplete assembly can trip it. A draft that simply does not contain the locus is indistinguishable from one that has lost it. In testing, a 49 contig neotomae draft 370 kb short of a full genome was flagged yes on that basis alone.

So read yes as "the O-antigen band is not where it should be, look closer", and do not read no as "smooth". What it does catch is shown in Validation.