Expand MAC doesn't expand XX codes

[lgragert]

XX codes are not considered valid MACs.

Should they be?

pyard --expand-mac "B*15:XX"
B*15:XX is an invalid MAC.

pyard --expand-mac "B*15:AB"
B*15:01/B*15:02

[lgragert]

Users of the NMDP MAC Service UI will find that it expands XX codes, so they might be confused that this function does not. I suppose we need the error message to point them at this page so they understand that XX codes are not MAC codes? https://glstring.org/xxcodes.html

[pbashyal-nmdp]

I'd think an equivalent expansion would be:

pyard --gl "B*15:XX" -r "P"

It's probably be a good idea to mention that XX codes are not MAC codes.

[lgragert]

Hmmmm... what features does --expand-mac have that --gl does not?

[pbashyal-nmdp]

--expand-mac just expands the MAC and validates each of the expansion makes sense.

--gl is paired with -r to reduce to certain level.
e.g.

pyard --gl "B*15:XX" -r "G"

[lgragert]

Is --expand-mac precisely equivalent to --gl -r "U2" then, or does it differ in some way?

[pbashyal-nmdp]

--expand-mac doesn't involve reduction.

e.g. AB MAC is defined as:

code  alleles
----  -------
AB    01/02

When you --expand-mac the MAC C*04:AB. It expands to C*04:01 and C*04:02 but only C*04:01 is a valid C allele. So it returns that as the expansion.

❯ pyard --expand-mac "C*04:AB"
C*04:01

versus A*04:AB

 pyard --expand-mac "A*04:AB"

produces no result because both expansion A*04:01 and A*04:02 are not valid alleles.

[lgragert]

Got it, so --expand-mac is like --gl -r "U2" except the latter will remove invalid alleles for a particular DB version.