Unambiguous patient genotype should be given match probability results

[benbelow]

Currently if a patient genotype is:

(a) Unambiguous with respect to p-groups (used for matching calculation) (b) Not represented in the selected (global) HF set

Then the patient HLA is marked as "unrepresented", and no result is given.

In practice, HF sets should not be needed for such patients - we know the p groups exactly, so the probability of this genotype is 1. (We don't know the phase / haplotypes involved, but we don't need phase information to calculate match probabilities).

We should change the algorithm such that patients unambiguous with respect to p-groups are not treated as unrepresented, and the donor's HLA only will impact the match probability.

The same will be true of donors (if less likely) - any donor unambiguous wrt p groups will be fully calculatable without needing to look up haplotype probabilities.

[benbelow]

example HLA:

"A": {
"Position1": "*01:01",
"Position2": "*23:01"
},
"B": {
"Position1": "*44:03",
"Position2": "*57:01"
},
"C": {
"Position1": "*04:09N",
"Position2": "*06:02"
},
"Dpb1": {
"Position1": "*112:01",
"Position2": "*114:01"
},
"Dqb1": {
"Position1": "*03:01",
"Position2": "*04:01"
},
"Drb1": {
"Position1": "*08:07",
"Position2": "*11:04"
}
}

[benbelow]

Testing by both @mmelchers and myself has confirmed that we're now seeing probabilities in such searches where we didn't previously. Closing this issue as completed.