Infer the timing of the events of interest fitting the appropriate TOSCA model

Usage

fit(
  x,
  n_iterations = 10000,
  n_chains = 4,
  warm_up = 5000,
  cores = 4,
  adapt_delta = 0.95,
  stepsize = 0.01,
  seed = 5,
  model_name = "Driver",
  dormancy = F,
  verbose = F,
  initialisation = F,
  init_fun = NULL,
  max_mrca = NA,
  reg_dormancy = 0
)

Arguments

x: TOSCA object
n_iterations: Number of iteration of each MCMC chain. The default is 10000.
n_chains: A positive integer specifying the number of Markov chains. The default is 4.
warm_up: A positive integer specifying the number of warmup (aka burnin) iterations per chain. The default is 5000.
cores: Number of cores to use when executing the chains in parallel
adapt_delta: target acceptance probability for the Metropolis step in HMC. The default is 0.95.
stepsize: how far the particle moves in each leapfrog step along the trajectory. The default is 0.01.
seed: seed of the computation. Default is 5.
model_name: Name of the model: "CNA" or "Driver"
dormancy: Boolean specifying if the model should include dormancy (available only for "CNA" model). The default is F
verbose: Boolen specifying if the fit function should output the iterations progression. The default is F.
initialisation: Boolean, default= F. If true, must specify an init_fun
init_fun: Vector with initial values used if initialisation = T.
max_mrca: maximun date for t_mrca
reg_dormancy: Boolean (0 or 1). If reg_dormancy = 0, a likelihood regularisation is imposed on the timing of dormancy end

Value

TOSCA object

Examples

library(TOSCA)
library(dplyr)
library(ggplot2)
data("exampleData_CNA")
set.seed(123)
mutations = exampleData_CNA$Mutations
parameters = exampleData_CNA$Parameters
samples = exampleData_CNA$Samples
therapies = exampleData_CNA$Therapies
x = init(mutations=mutations, samples=samples, therapies=therapies, parameters=parameters)
#> Error in init(mutations = mutations, samples = samples, therapies = therapies,     parameters = parameters): Input must be a data.frame.
fit = TOSCA::fit(x, model_name='CNA',
                 n_iterations = 10000,
                 n_chains = 4,
                 warm_up = 5000)
#> Error: object 'x' not found