HaveYouSwappedYourSamples

This project contains simple methods to measure sample relatedness and identify potential swaps and contamination

---

INSTALLATION

Clone the repository for the latest version: git clone http://github.com/PapenfussLab/HaveYouSwappedYourSamples.git There are some dependencies to the modules:

• HaveYouSwappedYourSamples.sh: None.
• concordance_script.py (calculate concordance module): Python.
• plotting_script.R (plot a heat map): The optional plotting routine depends on ggplot2 being installed within R.
• model_analysis.py (calculate Gaussian micture model): This optional module relies on Python and the Python libraries sklearn.mixture, numpy, and scipy.

---

DATA

The tool relies on the data being formatted and sorted in the following way:

1. each sample (and germ line) should have one tab delimited file that specifies
   SNP_ID CHR POS VAF
   The script actually only uses the VAF field (variant allele frequency)
2. the files specified above have to be lexicographically sorted by SNP_ID and SNP_IDs have to be unique within a sample. Specify the absolute path to each of the files in a text file (one row per file)
   Germ lines and samples should be specified in separate files.

---

USAGE

Run the "HaveYouSwappedYourSamples.sh" script followed by a module identifier:
"conc" "heat" "model" Each module needs optional and mandatory parameters:

1. conc -s (OPTIONAL): Symmetrical evaluation (consider homzygotes from samples and germ lines alike. By default only germ line SNPs are regarded : The file with paths to the samples : The file with paths to the gern lines (or samples, if an all vs. all concordance test is desired (use together with -s option).
2. heat : The file with the concordance measures (as produced by the "conc" module). : An output pdf file for the heat map to be saved as.
3. model : The file with the concordance measures (as produced by the "conc" module).