PRS-cs ancestry assignment seems wrong

[Gechang Yu]

Hi,

When I use the PennPRS to calculate trans-ancestry PRS by PRS-cs, I found a problem that even if I specified the ancestry of reference panel as East Asian, the PennPRS would always automatically specified it as European.

I tried twice but got the same error. The job IDs are 7ec72e5a-3355-4eb7-aa39-1432df27c8ea and eab805ec-5ac7-4c21-a66d-5439a7069aee

I checked the log file, which also showed that it specified the ancestry as EUR.

$methods
[1] "PRS-CS-auto"

$trait
[1] "T2DGGI_All_1KG_EAS_Penn"

$ancestry
[1] "EUR"

$LDrefpanel
[1] "1kg"

$phi
[1] 0.01

$coef_shrink
[1] 1

$allow_jump_sign
[1] TRUE

$use_MLE
[1] TRUE

$ensemble
[1] FALSE

Could you please help me check this issue? Thank you very much!

Gechang Yu

The Chinese University of Hong Kong

[Jin, Jin]

Hi Gechang,

 

Thank you for reporting the issue. Our team is working on it and will keep you posted.

 

Best,

Jin

 

From: pen...@googlegroups.com <pen...@googlegroups.com> on behalf of Gechang Yu <gecha...@gmail.com>
Date: Friday, December 5, 2025 at 4:51 AM
To: PennPRS <pen...@googlegroups.com>
Subject: [External] PRS-cs ancestry assignment seems wrong

You don't often get email from gecha...@gmail.com. Learn why this is important

--
您收到此邮件是因为您订阅了Google群组上的“PennPRS”群组。
要退订此群组并停止接收此群组的电子邮件，请发送电子邮件到pennprs+u...@googlegroups.com。
如需查看此讨论，请访问 https://groups.google.com/d/msgid/pennprs/68316ea4-d067-4ff5-bbad-1746df5ab55en%40googlegroups.com。
要查看更多选项，请访问https://groups.google.com/d/optout。

[Jin, Jin]

Hi Gechang,

 

We have updated our platform and the problem should be resolved now. Please let us know if you have more questions. Thank you for your patience!

 

Best,

Jin

 

如需查看此讨论，请访问 https://groups.google.com/d/msgid/pennprs/SN6PR04MB5072CFB4923456CBBA8DBFD1D1A4A%40SN6PR04MB5072.namprd04.prod.outlook.com。
要查看更多选项，请访问https://groups.google.com/d/optout。

[Gechang Yu]

Hi Jin,

Sorry for my late reply. Thank you very much for your help!

I found that the multi-ancestry PRS method list currently only has PROSPER. May I ask if you and colleagues plan to add PRS-CSx into the list? Thank you!

Best regards,

Gechang

[Gechang Yu]

In addition, I would like to report a bug for PRS-cs deployment in PennPRS. PennPRS provides two options for choosing fixed or auto Phi parameter in PRS-cs. But I checked the log file and found that when the user specify "auto" Phi parameter in PennPRS. The PennPRS will assign "NA" to Phi in PRS-cs. But this will lead to an error in PRS-cs:

File "/home/ubuntu/PRScs/PRScs.py", line 115, in <module>
    main()
  File "/home/ubuntu/PRScs/PRScs.py", line 92, in main
    param_dict = parse_param()
  File "/home/ubuntu/PRScs/PRScs.py", line 54, in parse_param
    elif opt == "--phi": param_dict['phi'] = float(arg)
ValueError: could not convert string to float: 'NA'

Therefore, I think that you should change this part of code in PennPRS. When the user chooses not to use a fixed Phi parameter, PennPRS should not specify the exact value for Phi parameter. In this case, PennPRS should delete the "Phi" parameter. I suggest to add an if/else statement in the code before specifying the parameters of PRS-cs. Thank you!

Best regards,

Gechang

在2025年12月16日星期二 UTC+8 06:02:03<Jin, Jin> 写道：