Table 1 Feature and requirement comparison of the cis methods used in this study

From: MR-link-2: pleiotropy robust cis Mendelian randomization validated in three independent reference datasets of causality

Method

Meta-analysis possible

Uses all SNPs in a region

Requires an LD reference

Robust to horizontal pleiotropy

Tests for

MR-link-2

Yes

Yes

Yes

Yes

Causality between two traits and presence of horizontal pleiotropy in the associated region

MR-IVW

Yes

No

Noa

No

Causality between two traits

MR-IVW-LD

Yes

No

Yes

No

Causality between two traits

MR-PCA

Yes

Yes

Yes

No

Causality between two traits

coloc

No

Yes

No

Yes

If a single causal SNP in a region is shared between two traits

coloc-SuSiE

No

Yes

Yes

Yes

If any of the causal genetic variants in a region are shared between two traits

  1. We compare 6 different methods for the identification of cis genetic regulation. We specify if meta-analysis of multiple regions is possible for each method, if the genetic method uses all genetic variants that are available in a genetic region, if the method requires an LD reference and if the method has been designed to handle horizontal pleiotropy. a MR-IVW needs LD information only to perform the pruning step.
Back to article page