Dear developers, we are running a big meta-analysis in which most studies are population-based, but also some family-based studies are included.
As more than 30 studies including biobanks are part of this effort many different software are being used, and meta-analyses will be carried out based on effect sizes and SE.
One of our family based studies is currently using FAMSCORE test in RVTEST but from the output we do not have effect sizes, only U and V statistics. As many studies did not use RVTEST, it is not an option to meta-analyse based on these statistics.
I was wondering if back conversions from Pvalues could be done?. As,
SE=1/sqrt(V.Stat)
And then based on that SE and PVALUE obtain the effect size for the alternative allele?
If this is not possible, could you maybe provide some insight in a possible solution.
Dear developers, we are running a big meta-analysis in which most studies are population-based, but also some family-based studies are included. As more than 30 studies including biobanks are part of this effort many different software are being used, and meta-analyses will be carried out based on effect sizes and SE.
One of our family based studies is currently using FAMSCORE test in RVTEST but from the output we do not have effect sizes, only U and V statistics. As many studies did not use RVTEST, it is not an option to meta-analyse based on these statistics.
I was wondering if back conversions from Pvalues could be done?. As, SE=1/sqrt(V.Stat) And then based on that SE and PVALUE obtain the effect size for the alternative allele?
If this is not possible, could you maybe provide some insight in a possible solution.
Thanks a lot. Carolina