Synopsis

The R/QTL Mapping Pipeline is a collection of scripts that streamline the process of building input files for Karl Broman's quantitative trait loci analysis package for R. The scripts connect to a database containing genotype data, filter markers using PLINK, and finally use this subset of markers along with some phenotype data to build csvsr-formatted input files for R/QTL.

Outline

make R/QTL phenotype input --> make PLINK inputs --> run PLINK --> make R/QTL genotype input --> perform R/QTL mapping

Usage

1. place a copy of run_pipeline.cmd in directory containing file with phenotype data

2. set parameters in run_pipeline.cmd

3. (Optional) if doing batch mapping with UW-Madison's Condor HTC cluster, place a copy of make_rdata.r in same directory as run_pipeline.cmd and specify what mapping jobs to do in make_rdata.r (see comments in make_rdata.r for details on specifying mapping jobs and map.r for general mapping information.)

4. execute run_pipeline.cmd in Windows Command Prompt* by simply typing

    run_pipeline.cmd
   

   * Note: Shift+Right-click inside a directory in File Explorer and select "Open command window here" to start Windows Command Prompt in current directory

For interactive mapping:

1. open interactive_mapping/[rqtl_mapping.r] (interactive_mapping/README.md) with RStudio or RGui
2. after loading data into a cross object, choose blocks of code to run

For batch mapping on UW-Madison Cluster:

1. see documentation

Summary of primary scripts

• run_pipeline.cmd
  • this is the backbone of the pipeline. It makes calls to scripts in the sub-directories and to make_rdata.r
  • produces directory in which all generated files (including files for R/QTL mapping) are placed
• make_rdata.r
  • outputs file for performing mapping on Condor HTC cluster
  • called by run_pipeline.cmd
• filter_markers/make_plink_inputs.py
  • outputs input files for PLINK which can then filter markers down to a sub-set that meet specified conditions (e.g. allele frequency, maximum missing rate)
  • called by run_pipeline.cmd
• make_rqtl_inputs/src/[make_rqtl_inputs.py] (make_rqtl_inputs/README.md)
  • outputs files with genotype and phenotype information in the csvsr format specified by R/QTL
  • called by run_pipeline.cmd
• [batch_mapping/*] (batch_mapping/README.md)
  • collection of scripts which perform R/QTL mapping on UW-Madison CHTC cluster
• interactive_mapping/[rqtl_mapping.r] (interactive_mapping/README.md)
  • R script with commonly used mapping commands, for use in R interactive session

Requirements

The following programs should be installed and exist in the Windows PATH environment variable

• [PLINK 1.9] (https://www.cog-genomics.org/plink2)
• [Python 3.X] (https://www.python.org/) (tested on Python 3.5)
• [R] (https://cran.r-project.org/) (tested on 3.2.4)

Required Python modules:

• [PYODBC] (https://mkleehammer.github.io/pyodbc/)

Install python modules from Windows Command Prompt via:

python -m pip install SomeModule