Open kkamii opened 2 years ago
I got the same issue. When I fitted the bslmm, it worked fine. But when I use the fitted bslmm to predict new phenotype, it gives me the error that "snp xxx has missing genotype for all individuals and will be ignored." for all snps and predict no value.
I carefully read the document. In section 4.9.3, it says
In particular, individuals with missing phenotypes will have predicted values, while individuals with non-missing phenotypes will have “NA”s.
That means even for prediction, you need to include the phenotype file. If the phenotype file contains 'NA', it gives predictions. Those non-“NA” value phenotypes will be used to analyze. But I still don't understand why it needs phenotype to do prediction.
Hi, I want to use bslmm model to predict my phenotype data.
I used - bslmm 1 option with -maf 0 -miss 1 -hwe 0 -r2 1 SNP QC options, so that all of snps can be analyzed
but, when I use -predict option, all of snps removed and predicted data have NA values.
how can I fix it?