pip install numpy pandas scipy

git clone https://github.com/DecodeGenetics/NCOurd.git 
cd ncourd

python bin/EMmodel.py --help

usage: EMmodel.py [-h] [--detfile DETFILE] [--ngroups NGROUPS1]
                  [--maxiter MAXITER] [--verbose] [--ntracts NTRACTS1]
                  [--nskip NSKIP] [--multiprocess] [--paramfile PARAMFILE]
                  [--method METHOD] [--dclass DCLASS] [--minmeth MINMETH]
                  [--maxspan MAXSPAN] [--nbinom_loc NBINOM_LOC]
                  [--randomsample RANDOMSAMPLE] [--accwts ACCWTS]
                  [--cubicgrid] [--twidx TWIDX] [--mink MINK]
                  tractfiles [tractfiles ...]

Run EM model on tractsets

positional arguments:
  tractfiles            tract weight function files to use

optional arguments:
  -h, --help            show this help message and exit
  --detfile DETFILE     file containing detection function
  --ngroups NGROUPS     number groups to fit - Default is #tractfiles
  --maxiter MAXITER     maximum number of iterations - Default is 1000
  --verbose             more printing
  --ntracts NTRACTS     number of tracts to take from each tractfile
  --nskip NSKIP         number of tracts to skip from each tractfile
  --multiprocess        use multiprocessing
  --paramfile PARAMFILE
                        File containing initial parameters for the EM model
  --method METHOD       Simultaneously "simul" or separately "sep" evaluate
                        shape and scale
  --dclass DCLASS       Class of distributions excluding the default negative
                        binomial distributions "nbinom" gamma "gamma" and
                        geometric "geom" distributions are supported
  --minmeth MINMETH     Method to pass to scipy.optimize.minimize for
                        simultaneous evaluation must support bounds
  --maxspan MAXSPAN     Discard tracts which are at least this long
  --nbinom_loc NBINOM_LOC
                        If nbiom distributions are used this sets location
                        parameter of nbiom distributions - Default is 1
  --randomsample RANDOMSAMPLE
                        Sample random tracts from input tracts with given seed
  --accwts ACCWTS       Accelerate weights for the first N iterations
  --cubicgrid           use cubic wgrid instead of linear for integration
  --twidx TWIDX         indexes to use in the tractfile
  --mink MINK           minimum shape parameter for the equivalent gamma
                        distributions (variance is at least mean**2*k)

python bin/Markers2DF.py --help

usage: Markers2DF.py [-h] [--regions REGIONS] [--dense] [--penetrance PENETRANCE] 
                     [--output OUTPUT] markerfile

Calculate detection function from informative markers writing the output to
[markerfile].p[PENETRANCE].detdf - The output file is suitable as input file for
EMmodel.py

positional arguments:
  markerfile            informative marker file

optional arguments:
  -h, --help            show this help message and exit
  --regions REGIONS     file with regions where the detecion functions 
                        should be calculated
  --dense               use denser grid
  --penetrance PENETRANCE
                        probability of NCO marker being gene converted
  --output OUTPUT       name of output file

python bin/tracts2twdet.py --help

usage: tracts2twdet.py [-h] [--penetrance PENETRANCE] tracts

calculate tract functions from tracts and writes them to the
file [tracts].p[PENETRANCE].twdet.gz - The output file is
suitable as input for EMmodel.py

positional arguments:
  tracts                file containing the tracts

optional arguments:
  -h, --help            show this help message and exit
  --penetrance PENETRANCE
                        probability of NCO marker becoming
                        gene converted