ping not working for DQB1*02:180

[mmaiers-nmdp]

DQB1*02:180 is in the DQB1*02:03P group but not the G group.

with ping: true this should reduce to DQB1*02:03 but it does not

Reduction Method: G
-------------------
DQB1*02:180:01G

Reduction Method: P
-------------------
DQB1*02:03P

Reduction Method: lg
--------------------
DQB1*02:180g

Reduction Method: lgx
---------------------
DQB1*02:180

Reduction Method: W
-------------------
DQB1*02:180:01/DQB1*02:180:02

Reduction Method: exon
----------------------
DQB1*02:180:01/DQB1*02:180:02

Reduction Method: U2
--------------------
DQB1*02:180

Reduction Method: S
-------------------
DQ2```

[mmaiers-nmdp]

I have confirmed that the necessary entry in the p_not_g table is missing and adding it manually results in the correct/expected behavior:

$ scripts/pyard -g "DQB1*02:180" -d /tmp/debug -r lgx
DRB1*02:03

on difference between this case and others that work (like C*06:17) is that the DQB1*02:180 is defined at 3-fields in the P-group file:
DQB1*02:180:01/DQB1*02:180:02

Perhaps 2-field was expected?

[pbashyal-nmdp]

I believe the lgx_group already contains the reduction so it doesn't add to the p_not_g table.

sqlite> select * from lgx_group where allele like 'DQB1*02:180%';
allele          lgx
--------------  -----------
DQB1*02:180     DQB1*02:180
DQB1*02:180:01  DQB1*02:180
DQB1*02:180:02  DQB1*02:180

And it only creates an entry in p_not_g table if it's not in lgx table.

[mmaiers-nmdp]

This is not the correct reduction if the mode is ping:True

This means for all 3 and 4-field alleles the ping logic is being bypassed.

in 3.61.0 there are 107 cases of ping exceptions

103 are 2-field alleles and all work correctly

2 are 3-field alleles and both are incorrect.
2 are 4-field alleles and both are incorrect.

This is a full list of offending cases:

C*03:404,C*03:04
DPB1*806:01,DPB1*04:01
DQB1*02:180,DQB1*02:03
DQB1*05:43,DQB1*05:03

The first two are 4-field alleles and the last two are 3-field alleles that are bypassing the ping logic