search

nmdp-bioinformatics / py-ard

HLA ARD Reduction in Python
https://py-ard.org/
GNU Lesser General Public License v3.0
17 stars 13 forks source link

Redux mode for first-field DNA specificities #327

Open lgragert opened 5 months ago

lgragert commented 5 months ago

py-ARD doesn't seem to have a redux mode for first-field DNA nomenclature yet.

The current US organ allocation system supports all three of these HLA specificities: DRB3*02:02, DRB3*02, and DR52, but py-ARD doesn't have a way to map between all of these categories yet. DRB3/4/5 alleles like DRB3*02:02 would have a rollup of DRB3*02 that is more specific than the serologic antigen rollup of DR52.

DQA1 and DPA1 loci don't have WHO-recognized serologic specificities, so this would be the only way to do a more ambiguous rollup for alleles such as DPA1*02:07.

Some GL strings with full-field HLA alleles will encompass several first-field specificities, which all should be provided in the output.

lgragert commented 4 months ago

Also noting that users that attempt an antigen-level redux for C*12 through C*18 alleles will return a null result from py-ARD. For example, for the allele specificity C*12:03, py-ARD does not have a more ambiguous redux category.

If we decide to support first-field DNA nomenclature rollup, should we provide an option to have the antigen redux for HLA-C also perform a first-field rollup for C*12 through C*18 alleles?

mmaiers-nmdp commented 1 week ago

What would be the preferred output for py-ard -r onefield -g "DPB1*104:01:1" should we also have py-ard -r twofield -g "DPB1*104:01:1"

lgragert commented 1 week ago

Probably the closest analogous thing for HLA-DPB1 alleles would be some rollup based on alleles sharing a DPB1 hypervariable region (HVR) amino acid motif or set of HVR motifs? https://pubmed.ncbi.nlm.nih.gov/32684412/