Error in rule ancestry_inference_i:

[peteryoon23]

Hi there,

Hello, I am currently studying PRS methodologies, and I really appreciate the development of GenoPred — it is a very useful and excellent tool. It aligns very well with my research topic, and I am eager to use it. However, I have been encountering persistent issues. I tried troubleshooting on my own, but I have not been able to resolve the problem, so I am reaching out for help.

I followed the installation instructions carefully and tested the pipeline using the provided test dataset. However, regardless of the method I use, I keep encountering the same error: “Error in rule ancestry_inference_i.”

As shown in the attached file, the issue seems to be that the ‘class’ package cannot be installed. Additionally, the ‘caret’ package also appears to be causing problems. Do you have any recommended solutions for this?

From my own troubleshooting, I suspect that there may be a version conflict between my R version and the ‘class’ and ‘caret’ packages. When I checked the analysis.yaml file, the r-base version is listed as 4.2.3. However, even when I upgrade or downgrade the r-base version, the same error still occurs.

Could you please advise on a good way to resolve this issue?

[opain]

Hi @peteryoon23, Thank you for the kind and clear question. This issue is likely caused by the R session within the Conda environment picking up R packages from your local user library (e.g., packages installed in your home directory) rather than those strictly defined in the Conda environment. Can I check you are specifying —use-conda when running the snakemake command? Also, if you are working on an HPC, ensure you haven’t loaded an R module that might be taking priority over the conda environment. One useful way to understand what is happening is to activate the conda environment that the rule is using (this will be printed on screen when it tries to run), and then go into R and check the library it is using by typing print(.libPaths()). You might be able to avoid the conflict by unloading R modules from your environment before running the pipeline, adapting your PATH variable, or by creating an .Renviron file containing R_LIBS_USER=/dev/null I am a little surprised this issue is only arising at the ancestry inference step, as these conflicts would normal cause issues further upstream, but this is what I would check first. Let me know if you are still stuck and I can try offer other solutions. Best wishes, Ollie

…

________________________________ From: peteryoon23 ***@***.***> Sent: Thursday, February 5, 2026 7:35:20 AM To: opain/GenoPred ***@***.***> Cc: Subscribed ***@***.***> Subject: [opain/GenoPred] Error in rule ancestry_inference_i: (Issue #165) [https://avatars.githubusercontent.com/u/253425464?s=20&v=4]peteryoon23 created an issue (opain/GenoPred#165)<#165> Hi there, Hello, I am currently studying PRS methodologies, and I really appreciate the development of GenoPred — it is a very useful and excellent tool. It aligns very well with my research topic, and I am eager to use it. However, I have been encountering persistent issues. I tried troubleshooting on my own, but I have not been able to resolve the problem, so I am reaching out for help. I followed the installation instructions carefully and tested the pipeline using the provided test dataset. However, regardless of the method I use, I keep encountering the same error: “Error in rule ancestry_inference_i.” As shown in the attached file, the issue seems to be that the ‘class’ package cannot be installed. Additionally, the ‘caret’ package also appears to be causing problems. Do you have any recommended solutions for this? 111.png (view on web)<https://github.com/user-attachments/assets/f3d64ab8-c103-4875-820f-2345f72c4a13> From my own troubleshooting, I suspect that there may be a version conflict between my R version and the ‘class’ and ‘caret’ packages. When I checked the analysis.yaml file, the r-base version is listed as 4.2.3. However, even when I upgrade or downgrade the r-base version, the same error still occurs. Could you please advise on a good way to resolve this issue? — Reply to this email directly, view it on GitHub<#165>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AKN4CSBGXJROYL4SFDPB35T4KLXDRAVCNFSM6AAAAACUB2OGSWVHI2DSMVQWIX3LMV43ASLTON2WKOZTHEYDAMJTGU4DGOA>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[peteryoon23]

Hi Ollie, Thank you for the kind explanation. Thank you so much. Now I'm finally getting a grasp of the issue. 1. I use —use-conda when running the snakemake command 2. Now I am using a server PC and not sure this PC is HPC or not 3. As you recommended, check the library with (.libPaths()) Here is the print of the library path. It looks like the R version conflict issue. I want to change the directory, but I'm hesitant to do so because this is a shared computer. How should I handle the path configuration? [image: image.png] thank you Best regards, Peter 2026년 2월 5일 (목) PM 5:16, Oliver Pain ***@***.***>님이 작성:

…

*opain* left a comment (opain/GenoPred#165) <#165 (comment)> Hi @peteryoon23, Thank you for the kind and clear question. This issue is likely caused by the R session within the Conda environment picking up R packages from your local user library (e.g., packages installed in your home directory) rather than those strictly defined in the Conda environment. Can I check you are specifying —use-conda when running the snakemake command? Also, if you are working on an HPC, ensure you haven’t loaded an R module that might be taking priority over the conda environment. One useful way to understand what is happening is to activate the conda environment that the rule is using (this will be printed on screen when it tries to run), and then go into R and check the library it is using by typing print(.libPaths()). I am a little surprise this issue is only arising at the ancestry inference step, as these conflicts would normal cause issues further upstream, but this is what I would check first. Let me know if you are still stuck and I can try offer other solutions. Best wishes, Ollie ________________________________ From: peteryoon23 ***@***.***> Sent: Thursday, February 5, 2026 7:35:20 AM To: opain/GenoPred ***@***.***> Cc: Subscribed ***@***.***> Subject: [opain/GenoPred] Error in rule ancestry_inference_i: (Issue #165) [https://avatars.githubusercontent.com/u/253425464?s=20&v=4]peteryoon23 created an issue (opain/GenoPred#165)< #165> Hi there, Hello, I am currently studying PRS methodologies, and I really appreciate the development of GenoPred — it is a very useful and excellent tool. It aligns very well with my research topic, and I am eager to use it. However, I have been encountering persistent issues. I tried troubleshooting on my own, but I have not been able to resolve the problem, so I am reaching out for help. I followed the installation instructions carefully and tested the pipeline using the provided test dataset. However, regardless of the method I use, I keep encountering the same error: “Error in rule ancestry_inference_i.” As shown in the attached file, the issue seems to be that the ‘class’ package cannot be installed. Additionally, the ‘caret’ package also appears to be causing problems. Do you have any recommended solutions for this? 111.png (view on web)< https://github.com/user-attachments/assets/f3d64ab8-c103-4875-820f-2345f72c4a13> From my own troubleshooting, I suspect that there may be a version conflict between my R version and the ‘class’ and ‘caret’ packages. When I checked the analysis.yaml file, the r-base version is listed as 4.2.3. However, even when I upgrade or downgrade the r-base version, the same error still occurs. Could you please advise on a good way to resolve this issue? — Reply to this email directly, view it on GitHub< #165>, or unsubscribe< https://github.com/notifications/unsubscribe-auth/AKN4CSBGXJROYL4SFDPB35T4KLXDRAVCNFSM6AAAAACUB2OGSWVHI2DSMVQWIX3LMV43ASLTON2WKOZTHEYDAMJTGU4DGOA>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***> — Reply to this email directly, view it on GitHub <#165 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/B4NPOOEQOMDFK4BJENIKAMD4KL36DAVCNFSM6AAAAACUB2OGSWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNJRG42TSOJZGQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[opain]

Hi Peter,

Glad things are making more sense.

Sorry, I can't see the output of .libPaths()? Can you send again? I will be able to give the clearest advice if you can upload screenshots of you

• trying to run the pipeline,
• then loading the conda environment the pipeline was trying to use,
• then opening R,
• and then running .libPaths().

This will show me the correct version of R is being loaded, and then show me which library the R packages are being read from.

If you think it is because the wrong R library is being used. Try running the following command just before running the snakemake command.

export R_LIBS_USER=/dev/null

If you are working on a shared server, and you are not sure what is going on, it might be easier to ask the admin for the server or someone local how handle PATH/conda related issues. I am very happy to try and help, but might take longer.

Best wishes,

Ollie

[peteryoon23]

Hi Ollie, Sorry for the output of .libPaths()'s screen shot I just have copy the output as below

print(.libPaths())

[1] "/home/bts/R/x86_64-pc-linux-gnu-library/4.2" [2] "/usr/local/lib/R/site-library" [3] "/home/bts/anaconda3/envs/imputation/lib/R/library" I would like to ask someone who knows how to handle PATH/conda related issues. However, the previous employee departed without a formal transition, and the current people are unfamiliar with the details. Thank you for kind help and plz have a look. I reckon that In case [3], The directory under the env folder should be 'genopred' rather than imputation... right? I don't mind if it takes a while. Best regards Peter 2026년 2월 9일 (월) PM 5:53, Oliver Pain ***@***.***>님이 작성:

…

*opain* left a comment (opain/GenoPred#165) <#165 (comment)> Hi Peter, Glad things are making more sense. Sorry, I can't see the output of .libPaths()? Can you send again? I will be able to give the clearest advice if you can upload screenshots of you - trying to run the pipeline, - then loading the conda environment the pipeline was trying to use, - then opening R, - and then running .libPaths(). This will show me the correct version of R is being loaded, and then show me which library the R packages are being read from. If you think it is because the wrong R library is being used. Try running the following command just before running the snakemake command. export R_LIBS_USER=/dev/null If you are working on a shared server, and you are not sure what is going on, it might be easier to ask the admin for the server or someone local how handle PATH/conda related issues. I am very happy to try and help, but might take longer. Best wishes, Ollie — Reply to this email directly, view it on GitHub <#165 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/B4NPOOFRXOTYNA7J52P3VD34LBDJZAVCNFSM6AAAAACUB2OGSWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNZQGI2TEMJXHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[opain]

Hi Peter,

OK. Happy to help. Thank you for sending the output of print(.libPaths()). It is strange as the ancestry_inference conda environment is not listed at all. This is likely because the conda environment wasn't activated before opening R.

Please can you send me screenshots I asked for, so I have a better understanding of what might be going wrong. Or email me (oliver.pain@kcl.ac.ukl), and we can arrange a call online to resolve the issue.

I would guess doing the following would resolve your issue: Make sure you are using the --use-conda flag when running the snakemake command. And then also, run export R_LIBS_USER=/dev/null before running the snakemake command.

Best wishes,

Ollie