hapflk

This archive contains hapmap3 based input files needed to run hapflk in a small (4Mb) region of the human genome harbouring the LCT gene.

The files contain dense genotyping data on 370 individuals from 4 populations (CEU,CHB,TSI) and 497 SNPs.

An R script for representing cluster frequencies from hapflk output.

This script requires the following R packages installed:

RColorBrewer
ggplot2
grid

Usage:

first chmod +x hapflk-clusterplot.R

./hapflk-clusterplot.R output_file.bz2
...

The two scripts attached allow to build and plot local population trees. The principle is

1. To recompute Reynolds distances between population in a region. This is done using the local_reynolds.py python (2.7) script. Requires numpy python package.
   ...

This python script estimates the hapflk chi-squared density and corresponding p-values.
The script requires the numpy, scipy and statsmodels python packages.

The input file must be a .hapflk result file, possibly containing all values from a genome-wide scan. The script creates a new file (suffix .hapflk_sc) which is the original file plus 2 new colums: the rescaled statistic and the corresponding p-values.
...

This R script allows to create input files to extract "causal" variants from an FLK eigen analysis as in Rochus et al. (2017), using the Caviar software