Defining sample analysis options per-locus should be available

[ressy]

Currently all sample analysis options are provided in the configuration file and applied for all loci, but there's an obvious use case for allowing per-locus values (e.g. varying propensity for PCR stutter in one locus over another). Allowing optional columns in the locus attributes table would be a straightforward way to handle this.