Getting Started with mrpipeline

mrpipeline provides a streamlined interface for Mendelian randomisation (MR) and colocalization analysis, with a focus on proteomic GWAS data from deCODE and 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)

Mendelian randomisation

mrpipeline ships with bundled test datasets for CD40 protein and Sjogren’s disease. Use these to explore the package without any external data.

# Bundled datasets: cd40_exposure (formatted exposure), sjogren_outcome (outcome)
bfile <- system.file("extdata", "ld_ref", package = "mrpipeline")

# Run cis-MR
mr_res <- run_mr(
  exposure = cd40_exposure,
  exposure_id = "CD40",
  outcome = sjogren_outcome,

  outcome_id = "SjD",
  instrument_region = list(chromosome = "20", start = 44746911, end = 44758502),
  bfile = bfile,
  methods = c("ivw", "egger", "weighted_median")
)

# Inspect results
mr_res
summary(mr_res)

# Plot (requires ggplot2)
plot(mr_res, type = "scatter")
plot(mr_res, type = "forest")

The mr_result object stores the full results table, harmonised instruments, F-statistics, Steiger filtering output, and any skipped methods — accessible via mr_res$results, mr_res$instruments, etc.

Colocalization

coloc_res <- run_coloc(
  exposure = cd40_exposure,
  outcome = sjogren_outcome,
  gene_chr = "20",
  gene_start = 44746911,
  gene_end = 44758502,
  bfile = bfile,
  methods = "abf"
)

coloc_res
summary(coloc_res)
plot(coloc_res, type = "pp_bar")

Gene coordinate lookup

Look up genomic coordinates for HGNC gene symbols via Ensembl (requires the biomaRt package):

coords <- get_gene_coords(c("CD40", "APOE"), build = "grch38")
coords

These coordinates can be passed directly to run_mr() and run_coloc() via the instrument_region and gene_* arguments.

Formatting exposure data

mrpipeline includes formatters for common proteomic GWAS sources:

format_pqtl_decode() — deCODE genetics
format_pqtl_ukbppp() — UKB-PPP (Olink)
format_single_cell_onek1k() — OneK1K single-cell eQTL

Each returns data formatted for TwoSampleMR, ready to pass to run_mr().