Heterogeneity test in a subgroup analysis for a Mendelian Randomization study with Multiple Genetic Variants

Question

[Fadoua]

Dear all, I am conducting an MR study, I need to check for heterogeneity after doing my subgroup analyses. I am using R software, meta-package, metareg function. my first question is an initial visual inspection for heterogeneity important? and my second question concerns the input of the metareg function. is it the causal estimates from each genetic variant Beta using the approximate standing errors SE and their values in my subgroups? and finally, even though MR analysis checks for heterogeneity, is it a must to recheck it after doing subgroup analysis? thank you