+A NOTE ON SIMULATION FILE STRUCTURE:
+The CMS pipeline right now is set up to expect simulations in the following formats:
+Neutral simulations are all in a single directory, and are named according to the following convention:
+ bestfit_neutral_params_[iteration number]
+When setting neutralprefix in the parameter files, there is no need to include the underscore between 'params' and [param number].
+Simulations with selection are in 90 different directories, named param_1 to param_90. Within each directory, there are 100 subdirectories, one for each iteration of the parameter set:
+Within each directory, output for each individual simulation is named according to the following convention:
+ selection_params_[param number]_[iteration number]
+When setting selectionprefix in the parameter files, there is no need to include the underscore between 'params' and [param number].
+If either of these change, path structures in cms_process_simulations_selection.pl and cms_process_simulations_neutral.pl wll have to be adjusted accordingly. However, because each cosi iteration yields multiple files the folder/subfolder structure is far more efficient than gathering all output in a single directory, and so maintaining it may be a wise idea.
+Additionally, because these simulations are demographically explicit, and following the Schaffner model, population 1 is European-like, population 2 is Asian-like and population 3 is African-like. This is hardcoded into some of the latter scripts in the pipeline, as is the ID of the selected population in the simulations with population-specific selection. If this changes, it will be necessary to edit some of the R scripts called by cms_compute_cms.pl accordingly.