inst/tests/README.md

pQTL MR (instrumental variables) and outcomes (MR-Base)

Experiment

File Instruments.csv was converted from its Instruments.xlsx and can be read with Ins <- read.csv("Instruments.csv"),Phenotype %in% c("ABO","LIFR")) and or directly from awk -FS="," 'NR==1||$1~/ABO/||$1~/LIFR/' Instruments.csv > Ins.csv.

source("pqtlMR.R")
Ins <- read.csv("Ins.csv")
ids <- c("ieu-a-7","ebi-a-GCST007432")
pqtlMR(Ins,ids)

The original script by Jie Zheng

library("readxl")
Ins <- data <- read_excel("Instruments.xlsx",1)
ids <- scan("outcome.id.txt",what="")

handles three files

  1. Instruments.xlsx
  2. outcome.id.txt
  3. MR-pQTL-script.R

A toy example

This offers possibility to simplify two-sample MR,

abc <- function(a=1,b=2,c=3) list(sum=a+b+c,prod=a*b*c)
test <- function(x,y=x$sum,z=x$prod,u=y+z) u
test(abc())

but unfortunately the scheme does not always work.

Legacy implementation

These can be seen from utils.R and pqtlMR.Rd and run_TwoSampleMR.Rd.



jinghuazhao/pQTLtools documentation built on Oct. 22, 2024, 12:56 a.m.