Hi there! Based on my understanding running mv_multiple with instrument_specific=TRUE flag runs MR using only instruments specific to each exposure. Is this any different then from running pairwise MR between each of the exposures with the given outcome? Does the mv_multiple function have any extra steps to account for potential collinearity or overlapping snp effects?
Thanks in advance!