iCARH (version 2.0.2.1)

iCARH.simulate: Simulates longitudinal data based on the iCARH model.

Description

Simulates longitudinal data based on the iCARH model. Returns two types of datasets with relevant parameters (see below).

Usage

iCARH.simulate(
  Tp,
  N,
  J,
  P,
  K,
  path.names = NULL,
  path.probs = FALSE,
  pathway.perturb.ratio = 0.5,
  Ygroupeff = NULL,
  Zgroupeff = NULL,
  fe = 0,
  num.corr.y = 0,
  beta.val = NULL,
  sigma2 = 1,
  arz = 0.7,
  sdx = 0.01
)

Arguments

Tp

number of time points

N

number of samples (by default first N/2 controls and last N/2 cases)

J

number of metabolites

P

number of pathways (will probably change)

K

number of bacteria profiles (Y variables)

path.names

pathways to sample from as specified in KEGG. If not specified, path.probs will be considered.

path.probs

if TRUE, KEGG like density of pathways per metabolite is used to sample from. If scalar, path.probs is the expected ratio of metabolites in each pathway. Needs to be specified if path.names is not.

pathway.perturb.ratio

expected ratio of perturbed pathways

Ygroupeff

vector of 2xK variables (treatment effect on Y variables)

Zgroupeff

vector of 2 variables for treatment effect

fe

fixed effect

num.corr.y

number of correlated Y variables. The last num.corr.y will be highly correlated to the first num.corr.y variables

beta.val

beta values (regression coefficients) to sample from. Values will be randomly sampled if not specified.

sigma2

individual variance of metabolites

arz

autoregressive coefficient for treatment simulation

sdx

noise for autoregressive process, recommended value is 0.01

Value

list with the following objects :

XX

metabolomics data, X data

Y

additional omic data, Y data

Z

treatment

beta

effects of Y variables on X variables, column K+1 represents effect of treatment on X variables

pathways

pathway adjacency matrices

path.perturb

which pathways are perturbed?

phi

"spatial" dependence parameter, indicative of pathway perturbation

arx

autoregressive coefficients for X data

ary

autoregressive coefficients for Y data

Examples

# NOT RUN {
 data.sim = iCARH.simulate(4, 8, 10, 2, 2, path.probs=0.3, Zgroupeff=c(0,4),
beta.val=c(1,-1,0.5, -0.5))

# }