search

zhilizheng / SBayesRC

GNU General Public License v3.0
25 stars 5 forks source link

effect allele frequency is missing in summary statistics #32

Closed SoleilChenxu closed 4 months ago

SoleilChenxu commented 4 months ago

Hi Zhili,

Thank you for your excellent work. I am applying the method in my analysis. I have got a question maybe you could give the suggestion.

The effect allele frequency is not available in my summary statistics, making it problematic to generate the cojo format file. You have suggested in previous converstion to use the frequency information in snp.info file. However, there are more than 10 million SNPs in my summary statistics. Some of the SNPs are not present in snp.info file. Do you have any advice regarding this? Can I still use the allele frequency in snp.info although the SNPs would be missing? Will this have an influence on the results?

Thanks in advance.

Best, Chenxu

zhilizheng commented 4 months ago

HI @SoleilChenxu ,

Yes, if you are using our LD resources, just keep the same set of SNPs. The other SNP not in common will still be removed in the QC steps, hence the allele frequency information is not used for those SNPs.

Regards, Zhili

SoleilChenxu commented 4 months ago

Hi Zhili,

Briliiant. I see. I do use the provided LD resources. Thank you!

Best, Chenxu

SoleilChenxu commented 4 months ago

HI @SoleilChenxu ,

Yes, if you are using our LD resources, just keep the same set of SNPs. The other SNP not in common will still be removed in the QC steps, hence the allele frequency information is not used for those SNPs.

Regards, Zhili

Hi Zhili,

Just a quick follow-up question. If I use the allele frequency in the provided resouces. I guess it's also better to use the sample size (N) in snp.info file right? Or I can use the N in the original summary statistics?

Best, Chenxu

zhilizheng commented 4 months ago

Hi @SoleilChenxu ,

No, you have to use your own per-SNP sample size. Only the allele frequency can be used from our LD panel, if you are sure your data quality is good.

Regards, Zhili

SoleilChenxu commented 4 months ago

Hi @SoleilChenxu ,

No, you have to use your own per-SNP sample size. Only the allele frequency can be used from our LD panel, if you are sure your data quality is good.

Regards, Zhili

Hi Zhili,

Okay I see. Thank you very much!

Best, Chenxu