chmod +x install.sh
./install.sh
ABC_WGD is made to investigate various models of speciation between a diploid and a tetraploid
Species 1 is considered as being diploid, species 2 is the tetraploid.
![model]
Prior distributions will feed the coalescent simulator msnsam, and are generated by priorgen. Priorgenwgd is a binary file from a compiled C++ code. It requires a bpfile and an argfile to work.
Values for all parameters are written in the file priorfile.txt, one line per multilocus simulations.
Summary statistics are directly computed from the msnsam's output. For each locus, mscalc will compute:
| Statistics | Description |
|---|---|
| bialsites | number of SNPs in the alignment |
| sf XY | number of fixed differences between species X and Y / locus length |
| sx X | number of exclusively polymorphic positions in species X / locus length |
| ss XY | number of shared biallelic positions between species X and Y/ locus length |
| pi X | Tajima’s Theta within species X |
| theta X | Watterson’s Theta witin species X |
| pearson_r_pi XY | correlation’s coefficient for pi over orthologs between X and Y |
| pearson_r_theta XY | correlation’s coefficient for theta over orthologs between X and Y |
| Dtaj X | Tajima’s D for species X |
| div XY | raw divergence Dxy measured between X and Y |
| netdiv XY | net divergence Da measured between X and Y |
| minDiv XY | smallest divergence measured between one individual from X and one from Y |
| maxDiv XY | highest divergence measured between one individual from X and one from Y |
| Gmin XY | minimum divergence between one sequence from X and one from Y minDivXY divided by the average divXY |
| FST XY | FST between X and Y compute as 1-(pi_X + pi_Z) / (2*pi_XY) |
An array of statistics corresponding to the average statistics computed over loci and their standard deviation will be returned every multilocus simulation and written in the file ABCstat.txt.
To run the simulations, simply use the following command:
run_ABC_polyploid.py [model] [argfile] [bpfile]
Ex: run_ABC_polyploid.py diso2_auto argfile_diso2_auto.txt bpfile
This example will run 10,000 multilocus simulations, with an autopolyploidization and a disomic inheritance.
run_ABC_polyploid.py will simply execute the pipeline
priorgen | msnsam | mscalc
The output files are ABCstat.txt (containing the computed summary statistics) and priorfile.txt (containing the parameter values used to simulate the data from which the summary statistics were calculated).
Model in [diso1, diso2_auto, diso2_allo, tetra1, tetra2_auto, tetra2_allo, hetero1, hetero2_auto, hetero2_allo]
This package requires:
pypy (has to be linked to the user's bin)
numpy
msnsam
Statistical comparison between "observation" and "simulations" can be made using various R libraries (abc, abcrf).