How to generate Enformer features for custom population (East Asian)?

[qq731936287]

I'm trying to use scPrediXcan with East Asian population data (504 individuals from 1000 Genomes).
I have the VCF files ready. Could you provide guidance on how to use the enformer_pipeline to
generate personalized epigenomic features for these individuals?

My setup:

• 504 East Asian individuals
• VCF files for all 22 chromosomes
• Reference genome (hg38) ready to download
• Gene annotations available

Thank you!

[hakyim]

Charles: could you share your code in github.com/hakyimlab/scpredixcan where you perform the prediction for the EU individuals in 1000G? FYI: this will cost you about 3000 GPU hours.

…

On Thu, Oct 9, 2025 at 11:02 AM kun2 ***@***.***> wrote: *qq731936287* created an issue (hakyimlab/scPrediXcan#8) <#8> I'm trying to use scPrediXcan with East Asian population data (504 individuals from 1000 Genomes). I have the VCF files ready. Could you provide guidance on how to use the enformer_pipeline to generate personalized epigenomic features for these individuals? My setup: - 504 East Asian individuals - VCF files for all 22 chromosomes - Reference genome (hg38) ready to download - Gene annotations available Thank you! — Reply to this email directly, view it on GitHub <#8>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAW2ROKSDHSKOVUS3GBGZSD3W2BK3AVCNFSM6AAAAACIX3JG4CVHI2DSMVQWIX3LMV43ASLTON2WKOZTGQ4TSNRXGQ2DINI> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[Yichao-Zhou]

The pipeline for running the personalized epigenomic features predictions is in the github repo. However, I can add my script example for prediction of European people. For further detailed technical questions, feel free to reach out to me via email, yichaozhou@uchicago.edu.

[hakyim]

Charles: Can you point to the specific files where the pipeline is? Brief instruction on how to use it for others without the need to email you personally?

…

On Fri, Oct 24, 2025 at 9:40 AM Yichao(Charles)Zhou < ***@***.***> wrote: *Yichao-Zhou* left a comment (hakyimlab/scPrediXcan#8) <#8 (comment)> The pipeline for running the personalized epigenomic features predictions is in the github repo. However, I can add my script example for prediction of European people. For further detailed technical questions, feel free to reach out to me via email, ***@***.*** — Reply to this email directly, view it on GitHub <#8 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAW2RONEVEWZJQC4OUHDBA33ZI26FAVCNFSM6AAAAACIX3JG4CVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTINBTGUYTSNZSGM> . You are receiving this because you commented.Message ID: ***@***.***>