improve handling of intrinsic alleles of fosA, oqxAB

[katholt]

Occasionally a chromosomal fosA sequence (which we don't screen for as they don't confer fosfomycin resistance) will be close enough (<10% divergent, e.g. 9.5%) to a mobile allele of fosA (which we do screen for as they confer fosfomycin resistance) that it gets detected and reported.
Need to fix this so that we screen for the chromosomal alleles so that they are accurately detected as best hit rather than a distant mobile allele; but don't report them if they are exact/very close matches to chromosomal alleles.
The same is probably true for oqxAB, need to review the sequences and decide if we can improve handling in this way also.
Note that SHV is dealt with separately now, with alleles recently curated (see Lam et al 2021 Kleborate preprint) and spectrum-changing mutations specifically tracked.