Paper: Identification and correction of sample mix-ups in expression genetic data

The full manuscript (with supplementary tables and figures) is here.

The paper is available at arXiv and as a formal journal article at G3:

Broman KW, Keller MP, Broman AT, Kendziorski C, Yandell BS, Sen Ś, Attie AD (2015) Identification and correction of sample mix-ups in expression genetic data: A case study. G3 5:2177-2186

The data are available at the Mouse Phenome Database, though not in exactly the form used in this repository.

The primary manuscript files are samplemixups_nolegends.Rnw and samplemixups_supp_nolegends.tex.

The Perl script add_legends.pl adds all of the legends, and then the .Rnw file is run through knitr to create a LaTeX file, and the two LaTeX files are sent through pdflatex and xelatex, respectively, to create PDFs.

Things are a bit tricky. In principle, the Makefile tells the full story, but the Analysis/R subdirectory has an asciidoc file for the analyses in the work. That directory has its own Makefile. Cached intermediate results are available at figshare: samplemixups_rcache.zip (This contains a bunch of .RData files that go in Analysis/R/Rcache.)

To compile everything, you can:

1. Download the cached intermediate results, samplemixups_rcache.zip and unzip them. This will populate Analysis/R/Rcache.

2. In Analysis/R, run

   R CMD BATCH grab_data.R

   This will download the primary data files. It's quite slow, as it's 2 GB of data to download.

3. In the primary directory, run make.

---

Necessary tools

• R
• Perl
• Python 2.7
• GNU make
• Asciidoc
• R packages: knitr, qtl, broman, lineup, ascii, data.table, igraph, beeswarm, RColorBrewer

---

To Do

• Do clean tests, with and without the intermediate files

---

License

The content in this repository is licensed under CC BY.