demonanalysis

Analysis and plotting of demon data

From a single simulation

To plot charts of variant allele frequencies and genotype sizes use plot_all_charts. For normal tissue, you should specify a value for max_allele_count so that the axes are appropriately scaled.

To plot Muller plots and grids use plot_all_images.

From a batch of simulations run on a cluster

Updating

To pull the latest github version into a cluster folder (using terminal):

cd XXX # replace XXX with the path to the cluster folder
git pull https://github.com/robjohnnoble/demonanalysis

To install that version on a cluster profile (using R):

library(devtools)
install("XXX/demonanalysis") # replace XXX with the path to the cluster folder

Set up

library(demonanalysis)
subfolder_name<-"XXX"# insert name of top-level folderinput_dir<- paste0("all_results/", subfolder_name) # folder containing results of the batchnum_parameters<- count_parameters(input_dir) # number of simulation parameters (first columns in data)output_dir_plots<- paste0("plots/", subfolder_name) # folder to receive image filesoutput_dir_data<- paste0("data/", subfolder_name) # folder containing data files

Check which simulations completed successfully

all_statuses(input_dir, summary=TRUE) # should be "Exit code 0" (when finished) or "So far no status" (while running)

Create plots for a batch of simulations

For growing tumours, to plot charts of variant allele frequencies and genotype sizes:

create_plots_batch(input_dir, output_dir=output_dir_plots, type="chart")

For normal tissue, to plot charts of variant allele frequencies and genotype sizes:

create_plots_batch(input_dir, output_dir=output_dir_plots, type="chart", max_size=50)

To plot Muller plots and grids:

create_plots_batch(input_dir, output_dir=output_dir_plots, type="plot")

Get general-purpose data

data<- all_output(input_dir, include_diversities=FALSE) # combined data for a batch of simulations, excluding diversity columns

Plot relationships between variables (using dataframe)

plot_curves_faceted(data, num_parameters, x_var="Generation", y_var="MeanBirthRate", output_filename="curves", output_dir=output_dir_plots)
# change x_var, y_var and output_filename as appropriate

Get additional data for forecasting

data<- all_output(input_dir) # combined data for a batch of simulations, including diversity columnsdata<- add_relative_time(data, start_size=5500, num_parameters=num_parameters) # add columns useful for plotting trajectoriesstart_size_range<-500*2^(0:8) # NumCells at time of initial measurement for forecastinggap_range<- (1:10)/10# gap between time of initial measurement and second measurementfinal_size<-1E5# waiting time is measured until tumour reaches this NumCells valuesummary<- get_summary(data, start_size_range, gap_range, final_size, num_parameters=num_parameters) # summary data for each simulation, for each combination of gap and final_sizecor_summary<- get_cor_summary(summary, c("DriverDiversity", "DriverEdgeDiversity"), num_parameters=num_parameters, min_count=5) # summary dataframe of correlations with "outcome", including all cellswait_cor_summary<- get_wait_cor_summary(summary, c("DriverDiversity", "DriverEdgeDiversity"), num_parameters=num_parameters, min_count=5) # summary dataframe of correlations with "waiting_time", including all cellsdepth_wait_cor_summary<- get_wait_cor_summary(summary, c(paste0("DriverDiversityFrom1SamplesAtDepth", 0:10), paste0("DriverDiversityFrom4SamplesAtDepth", 0:10)), num_parameters, min_count=5) # summary dataframe of correlations with "waiting_time" for different biopsy protocols