Skip to contents

mrpipeline

mrpipeline provides a streamlined interface for Mendelian randomisation (MR) and colocalization analysis, with a focus on proteomic GWAS data (deCODE, UKB-PPP). It wraps TwoSampleMR, coloc, and MendelianRandomization into a consistent workflow with S3 result objects and built-in sensitivity analyses.

Installation

Install from GitHub:

# install.packages("pak")
pak::pak("BZuckerman97/mrpipeline")

Quick start 

library(mrpipeline)

# Format exposure data (e.g. UKB-PPP pQTL)
exposure <- format_pqtl_ukbppp(cd40_sumstats, exposure_id = "CD40")

# Look up gene coordinates
coords <- get_gene_coords("CD40", build = "grch38")

# Run cis-MR
mr_res <- run_mr(
  exposure = exposure,
  exposure_id = "CD40",
  outcome = sjogren_outcome,
  outcome_id = "SjD",
  instrument_region = list(
    chromosome = coords$chromosome,
    start = coords$start,
    end = coords$end
  )
)
mr_res
summary(mr_res)

# Run colocalization
coloc_res <- run_coloc(
  exposure = exposure,
  outcome = sjogren_outcome,
  gene_chr = coords$chromosome,
  gene_start = coords$start,
  gene_end = coords$end,
  bfile = "path/to/ld_reference"
)
coloc_res
summary(coloc_res)

Vignettes