search

choishingwan / PRSice

A software package for calculating, applying, evaluating and plotting the results of polygenic risk scores
http://prsice.info
GNU General Public License v3.0
187 stars 89 forks source link

About the covariate file #361

Open acca1562 opened 3 months ago

acca1562 commented 3 months ago

Thank you very much for creating such an excellent product. As a beginner, I am sorry to trouble you with a question regarding the creation of PRS using PRSice2. My target data is divided by chromosome, and according to the PRSice2 homepage, it is possible to handle this by adding the code --target chr#. However, I am unsure how to handle the covariate file used when creating the final PRS, as the eigenvec file is also divided by chromosome. I believe that the covariate file needs to be a single file, but at present, I have ended up with separate covariate files for each chromosome. I would greatly appreciate any advice on how to address this issue. Thank you very much.

choishingwan commented 3 months ago

If your eigenvec is divided by chromosome, then you have not constructed it correctly. An eigenvec should be constructed across all chromosome. I’d suggest you to go back and study the basic of genome wide association study

Sam

On Mon, Jul 29, 2024 at 9:07 PM acca1562 @.***> wrote:

Thank you very much for creating such an excellent product. As a beginner, I am sorry to trouble you with a question regarding the creation of PRS using PRSice2. My target data is divided by chromosome, and according to the PRSice2 homepage, it is possible to handle this by adding the code --target chr#. However, I am unsure how to handle the covariate file used when creating the final PRS, as the eigenvec file is also divided by chromosome. I believe that the covariate file needs to be a single file, but at present, I have ended up with separate covariate files for each chromosome. I would greatly appreciate any advice on how to address this issue. Thank you very much.

— Reply to this email directly, view it on GitHub https://github.com/choishingwan/PRSice/issues/361, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAJTRYWI7FYGSIC33XZX4TLZO3RLTAVCNFSM6AAAAABLVKOPASVHI2DSMVQWIX3LMV43ASLTON2WKOZSGQZTMNRXGQ2TSNI . You are receiving this because you are subscribed to this thread.Message ID: @.***>

acca1562 commented 3 months ago

Thank you very much for your reply. I have one more question: If my target data is divided by chromosome after imputation, should I create the eigenvec file using the data that is not divided by chromosome before imputation, or should I merge the files after imputation and then create the eigenvec file? As I have just started studying genetic statistics, I intend to thoroughly learn the basis of GWAS from now on.

choishingwan commented 3 months ago

You don’t need to generate the eigenvec with imputed data. In fact, you should prune your data before calculating the pcs. You can refer to online tutorials for that

Sam

On Mon, Jul 29, 2024 at 9:52 PM acca1562 @.***> wrote:

Thank you very much for your reply. I have one more question: If my Target data is divided by chromosome after imputation, should I create the eigenvec file using the data that is not divided by chromosome before imputation? I apologize because this is a basic question. As I have just started studying genetic statistics, I intend to thoroughly learn the basis of GWAS from now on.

— Reply to this email directly, view it on GitHub https://github.com/choishingwan/PRSice/issues/361#issuecomment-2257308174, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAJTRYQYXK2QHJL3A67YZETZO3WWDAVCNFSM6AAAAABLVKOPASVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENJXGMYDQMJXGQ . You are receiving this because you commented.Message ID: @.***>

acca1562 commented 3 months ago

Thank you very much for kindly teaching me so much. I will make sure to study hard.