We found a strong spearman correlation of 0.21 ( P = 1.5 × 10 −235 ) between nominally significant estimates of h ^ Y and the MR-IVW absolute deviation from the weighted estimate (Fig. 5c ) (Supplementary Data 13 ), suggesting that MR-link-2 detects region-specific pleiotropy accurately (without the knowledge of estimates from other regions).
← all excerpts
MR-link-2: pleiotropy robust cis Mendelian randomization validated in three independent reference datasets of causality.
1
—
—