Hi,
So I've been using SAIGE for a while, and while performing the GBAT (gene based association tests), I would always set LOCO=False, because setting LOCO=True would cause some sort of parsing issue where the script would stdout "chunk is empty". For context, my genotypic data is in the PLINK format. Recently, based on some research I read, it seems like when working with a very large sample space which has high chances of having a significant amount of related individuals, having LOCO set to false would lead to a reduction in power. But how am I supposed to have LOCO set to true if it isn't able to parse the input data appropriately? Does LOCO=True only work when the input genotypic data is of VCF/BGEN format? What even is the point of having the option for LOCO if it doesn't even seem to work for the association tests? PS - I have set the chr=1 flag as well
Hi, So I've been using SAIGE for a while, and while performing the GBAT (gene based association tests), I would always set LOCO=False, because setting LOCO=True would cause some sort of parsing issue where the script would stdout "chunk is empty". For context, my genotypic data is in the PLINK format. Recently, based on some research I read, it seems like when working with a very large sample space which has high chances of having a significant amount of related individuals, having LOCO set to false would lead to a reduction in power. But how am I supposed to have LOCO set to true if it isn't able to parse the input data appropriately? Does LOCO=True only work when the input genotypic data is of VCF/BGEN format? What even is the point of having the option for LOCO if it doesn't even seem to work for the association tests? PS - I have set the chr=1 flag as well