fitOneTheta: Fit the maximum likelihood estimate of the parameters for a...

In Jmendo12/evemodel: Expression Variance Evolution Model

Description

Fit the maximum likelihood estimate of the parameters for a specific EVE model

Usage

fitSharedBeta(sharedBeta, tree, gene.data,
  colSpecies = colnames(gene.data), extra.var = NULL,
  lowerBound = c(theta = -99, sigma2 = 1e-04, alpha = 0.001),
  upperBound = c(theta = 99, sigma2 = 9999, alpha = 999),
  logTransPars = c("alpha", "sigma2", "beta"), cores = 1, fork = F)

fitOneTheta(tree, gene.data, colSpecies = colnames(gene.data),
  extra.var = NULL, lowerBound = c(theta = -99, sigma2 = 1e-04, alpha =
  0.001, beta = 0.001), upperBound = c(theta = 99, sigma2 = 9999, alpha =
  999, beta = 99), logTransPars = c("alpha", "sigma2", "beta"),
  cores = 1, fork = F)

fitTwoTheta(tree, gene.data, isTheta2edge,
  colSpecies = colnames(gene.data), extra.var = NULL,
  lowerBound = c(theta1 = -99, theta2 = -99, sigma2 = 1e-04, alpha =
  0.001, beta = 0.001), upperBound = c(theta1 = 99, theta2 = 99, sigma2 =
  9999, alpha = 999, beta = 99), logTransPars = c("alpha", "sigma2",
  "beta"), cores = 1, fork = F)

Arguments

sharedBeta	the shared beta parameter
tree	Phylogeny
gene.data	A matrix of expression values with samples in columns and genes in rows
colSpecies	A character vector with same length as columns in returned expression matrix, specifying the species, i.e. tip labels in the phylogeny, for the corresponding column.
extra.var	An optional matrix of technical variance for each expression value. Same dimensions as gene.data (default=NULL)
lowerBound	A named numeric vector of the lower bound for the estimated parameters
upperBound	A named numeric vector of the upper bound for the estimated parameters
logTransPars	A character vector with the names of the parameters that will be log transformed
cores	Number of parallel processes to run
fork	Use forking for parallel execution
isTheta2edge	Logical vector with same length as number of edges in tree specifying whether the corresponding edge theta parameter should be theta2 (TRUE) or theta1 (FALSE)

Value

A list with:

par

A matrix with the parameter estimates

ll

The log likelihood for each estimate

iterations

Number of iterations of the optimisation routine before reaching the estimate

convergence

Error code from the optim function. 0 means convergence.

message

Error message from the optim function.

Functions

• fitSharedBeta: Fit model with a given shared beta for all genes

• fitTwoTheta: Fit model with two different thetas assigned to specific edges

Jmendo12/evemodel documentation built on June 17, 2020, 8:56 p.m.