Aldy v4.4 incorrectly calls GeT-RM sample NA21781 that was correctly called by Aldy v2.2.6

[VasilisFlouris]

Hello there,

while testing Aldy v4.4 on CYP2D6-calling on samples from GeT-RM I encountered something unexpected. On a publication that used Aldy v2.2.6, sample NA21781 is correctly genotyped as *2x2/*68+*4 (supplementary materials), however, Aldy v4.4 that I used calls this sample as *4/*63+*65.

The BAM file for this sample was downloaded from the ENA website and used as is.

Aldy was used as :

aldy genotype -p wgs -g cyp2d6 path_to_file.bam

and the output was:

🐿  Aldy v4.4 (Python 3.10.6 on Linux 5.15.0-58-generic-x86_64-with-glibc2.35)
   (c) 2016-2023 Aldy Authors. All rights reserved.
   Free for non-commercial/academic use only.
Genotyping sample NA21781.bam...
Potential CYP2D6 gene structures for NA21781:
   1: 2x*1,1x*141.1001 (confidence: 100%)
   2: 2x*1,1x*61 (confidence: 100%)
   3: 2x*1,1x*63 (confidence: 100%)
Potential major CYP2D6 star-alleles for NA21781:
   1: 1x*2, 1x*4.021.ALDY, 1x*65 (confidence: 100%)
   2: 1x*4.021, 1x*63, 1x*65 (confidence: 100%)
Best CYP2D6 star-alleles for NA21781:
   1: *4.021 / *63 + *65 (confidence=100%)
      Minor alleles: *4.021, *63.001, *(65.001 +rs28371701 +rs28735595)
CYP2D6 results:
  - *4.021 / *63 + *65
    Minor: [*4.021] / [*63.001] + [*65.001 +rs28371701 +rs28735595]
    Legacy notation: [*4.021] / [*63] + [*65 +rs28371701 +rs28735595]
    Estimated activity for *65: uncertain function (evidence: L); see https://www.pharmvar.org/haplotype/187 for details
    Estimated activity for *4.021: no function (evidence: D); see https://www.pharmvar.org/haplotype/652 for details
    Estimated activity for *63: unknown

Could this be due to an error on my end?

Thank you.