Perform Mendelian Randomization based on the core gene hypothesis for polygenic exposures

This function performs Mendelian Randomization (MR) analyses based on the core gene hypothesis for polygenic exposures. MR Corge identifies a small number of putative core instruments that are more likely to affect genes with a direct biological role in an exposure and obtains causal effect estimates based on these instruments, thereby reducing the risk of horizontal pleiotropy.

Usage

mrcorge(
  harmonized_data,
  rank = "beta",
  K = 5,
  method_list = c("mr_ivw", "mr_weighted_mode", "mr_weighted_median"),
  seed = 1
)

Arguments

harmonized_data: Data frame containing the harmonized data from the TwoSampleMR package.
rank: Character, method to rank instruments, one of 'beta' (absolute per-allele effect size), 'h2' (per-variant heritability), 'pval' (significance), or 'zscore' (sample size-normalized z-score). Default is 'beta'.
K: Integer, number of instrument groups. Default is 5.
method_list: Character vector, list of MR methods to apply. Default is c("mr_ivw", "mr_weighted_mode", "mr_weighted_median").
seed: Integer, seed for random number generation. Default is 1.

Value

A list containing:

mrcorge: The main MR results.
result: All MR results including group-specific and cumulative estimates.
instrument: Data frame of instruments with their respective groups.
Q: Q statistics for heterogeneity.
I2: I2 statistics for heterogeneity.
het_P: P-values for heterogeneity tests.

Examples

# Example usage:
library(MRCorge)
data("HDL_CAD")
res <- mrcorge(HDL_CAD, rank='beta', K=20, method_list = c("mr_ivw"))
#> TwoSampleMR version 0.5.10 
#> [>] New: Option to use non-European LD reference panels for clumping etc
#> [>] Some studies temporarily quarantined to verify effect allele
#> [>] See news(package='TwoSampleMR') and https://gwas.mrcieu.ac.uk for further details
#> 
#> Warning:
#> You are running an old version of the TwoSampleMR package.
#> This version:   0.5.10
#> Latest version: 0.6.8
#> Please consider updating using remotes::install_github('MRCIEU/TwoSampleMR')
#> Analysing '0GbyEu' on '7n4jWF'
#> MR CORGE estimates: 
#>   id.exposure id.outcome outcome exposure                    method nsnp
#> 1      0GbyEu     7n4jWF outcome exposure Inverse variance weighted    6
#>             b        se      pval    Fstat group       category
#> 1 -0.03943587 0.1004654 0.6946652 750.7501     1 Group-specific
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 2 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 3 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 4 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 5 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 6 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 7 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 8 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 9 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 10 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 11 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 12 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 13 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 14 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 15 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 16 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 17 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 18 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 19 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining group-specific estimates for instrument group 20 / 20 ... 
#> obtaining cumulative estimates for instrument group 1 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 2 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 3 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 4 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 5 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 6 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 7 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 8 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 9 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 10 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 11 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 12 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 13 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 14 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 15 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 16 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 17 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 18 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 19 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'
#> obtaining cumulative estimates for instrument group 20 / 20 ... 
#> Analysing '0GbyEu' on '7n4jWF'