SA.wrapper.r: R wrapper to calculate statistics (F_st, Pi_x, Pi_y) using ABC (approximate bayesian)
Implicit Input parameter file: 'my.input.'
Input file format:
the first is just n and is used by the ABC package, the recombination rate, the frequency of the SA allele on the X chromosome, the frequency of the SA allele on the Y, SAS location (in rho) from the SDR (in real world unknown)
Example Input files:
A typical neutral simulation will therefore start with an input file (say input1) like: 1 100 0 0 125
The only real difference with simulations with selection is that lines 3 and 4 will now have different numbers (let's say in input2 this time): 1 100 0.1 0.9 125
Note: Varying strengths of selection (so differences in frequency on X vs Y) as well as varying locations (final line) should be considered. For now let's not worry about recombination rate and keep that fixed on 100 rho/Mb.
The strength of SAS is determined by the third and fourth parameters of the simulation - the frequency of the SA site on the X and Y respectively. When these frequencies are very different (say 0.9 vs 0.1), then there is strong SAS selection. When they are both 0 there is none, and when they are equal to each other there is balancing selection (related to SAS, but distinct).
Explicit Input parameter:
Explicit Output file: "outputn" with summary statistics
Usage: Rscript SA.wrapper.R
Usage example: Rscript SA.wrapper.r 1
SaSimSnps: precompiled simulation binary code (called by the SA wrapper)