Question: does PANTHER require unique SNP IDs in target .bim / annotation files?

[q1kx]

Hi, thank you for developing X-Wing.

I have a question about the input requirements for PANTHER.

Does PANTHER require SNP IDs to be unique in the target .bim file and annotation files?

I encountered a reproducible error in gibbs_sampler.py:

ValueError: cannot reshape array of size 90900 into shape (90893,1)

After checking my inputs, I found that:

• annot_EAS.txt has unique SNP IDs
• the target .bim file contains duplicated SNP IDs
• annot_EUR.txt also contains duplicated SNP IDs, apparently inherited from the target .bim

So I would like to confirm:

1. Are duplicated SNP IDs allowed in the target .bim?
2. Should annotation files be deduplicated by SNP ID before running PANTHER?
3. Could duplicated SNP IDs cause this reshape error?

Thanks very much for your help.