incorrect placement

[rvita]

HLA-A*25:03 protein complex & HLA-C*02:10 protein complex are showing up in the incorrect branch of MHC class II protein complex.
HLA-DQA1*01:02/DRB1*15:01 protein complex should be a child of HLA-DRB1*15:01 protein complex.
HLA-DRA*01:01/DRB1*03:01 protein complex should be a child of HLA-DRB1*03:01 protein complex

[jamesaoverton]

1. HLA-A*25:03 protein complex & HLA-C*02:10 protein complex were asserted under class II -- I will change to class I
2. HLA-DQA1*01:02/DRB1*15:01 protein complex this is mixing DQ and DR, which should it be?
3. HLA-DRA*01:01/DRB1*03:01 protein complex not sure...

[rvita]

1. HLA-A25:03 protein complex & HLA-C02:10 protein complex were asserted under class II -- I will change to class I
   OK
2. HLA-DQA101:02/DRB115:01 protein complex this is mixing DQ and DR, which should it be?
   YES, I checked the paper, it is stupid, but true.
3. HLA-DRA01:01/DRB103:01 protein complex not sure...
   not sure why it is not or if it should be?

[jamesaoverton]

2. That messes up the logic. We define HLA-DRB1*15:01 protein complex as "has part some HLA-DRA chain" not some HLA-DQA chain. Do you want to make a big change in the logic?

3. HLA-DRA*01:01/DRB1*03:01 protein complex says "serotype member of some HLA-DR3 serotype" but HLA-DRB1*03:01 protein complex says "serotype member of some HLA-DR17 serotype": the logic doesn't work the way you're asking if the serotypes don't match. It looks like there are other cases like this.

[rvita]

question 3 has been sent to John Sidney.
question 2: HLA-DRB115:01 protein complex should be defined as any complex that has a HLA-DRB115:01 chain, it should not depend on its alpha chain. DRB1*15:01 is a beta chain.

[jamesaoverton]

We talked to Bjoern today and decided:

1. straightforward fix
2. HLA-DQA1*01:02/DRB1*15:01 protein complex should go under a new "mixed" node
3. should be fixed by reworking the logic
4. we should rework the serotypes more generally