This archive contains code to perform the likelihood calculations described in Terhorst, Schlötterer & Song (2015).
There are two dependencies required to build this package which are not standard on all UNIX-like systems:
-
A C++11-compliant compiler (such as GCC 4.8 or higher). The clang compiler shipped with recent versions of Mac OS X should also work.
-
The google-sparsehash library (https://code.google.com/p/sparsehash/).
To compile the package, type make in the directory containing this
README file.
The file example.py is example of how to use this library to perform
inference. It analyzes a test data set (also included) in order to
test for and estimate selection. Refer to the comments in that file
for further details. Execute example.py using the wrapper script
example.sh in order to ensure that the threelocus library can be
located on your system's library path.