Page 1 of 1

NA and 0 as a values for beta_SNP

Posted: Tue Mar 08, 2016 9:43 am
by eliana
Hello!

i am using PROBABEL to run a GWAS. The output file shows either NaN or 0 as a Beta values. In a previous forum, this issue was presented for a few SNPs and running a regression in R was suggested. However, I am having the problem for all SNPs. What could be wrong and how I could fix it?. Many thanks in advance

Best,

Eliana

Re: NA and 0 as a values for beta_SNP

Posted: Tue Mar 08, 2016 9:52 am
by lckarssen
What kind of regression are you running? Linear, logistic or CoxPH?
What version of ProbABEL do you use?

Could it be that all your SNPs are rare/have a very low MAF? Please also check the imputation Rsq (both can be found in the info file).