MedZIM_func: Perform MedZIM model
In quranwu/MedZIM: The package performs the MedZIM model

Description Usage Arguments Details Value Examples

View source: R/MedZIM_func.R

The Outcome model is:

Y=β0+β1M+β2 1_{(M>0)}+β3X+β4X1_{(M>0)}+β5XM+ε

MedZIM_func(
  dat,
  xVar,
  yVar,
  taxon_name,
  libSize_name,
  obs_gt_0 = 2,
  obs_eq_0 = 2,
  inter_x_mg0 = T,
  inter_x_m = F,
  eval.max = 200,
  iter.max = 200,
  x_from = 0,
  x_to = 1,
  type1error = 0.05,
  paraJobs = 4
)

`dat`	The dataset used for the model.
`xVar`	The name of X (independent variable) in the data.
`yVar`	The name of Y (outcome variable) in the data.
`taxon_name`	The string that can identify the taxon variables.
`libSize_name`	The name of library size variable in the data.
`obs_gt_0`	The at least number of non-zero observations that a taxon should have to be included into the analysis.
`obs_eq_0`	The at least number of zero observations that a taxon should have to be included in group 1.
`inter_x_mg0`	Whether to include the β4 term into the model.
`inter_x_m`	Whether to include the β5 term into the model.
`eval.max`	Maximum number of evaluations of objective functions allowed in nlminb function.
`iter.max`	Maximum number of iterations allowed in nlminb function.
`x_from`	The value X starts from in evaluating the effect of X.
`x_to`	The value X ends to in evaluating the effect of X.
`type1error`	The type I error.
`paraJobs`	The number of cores for parallel computing.

The output list contains information of all taxons to be analyzed for mediation effects. Each taxon result is also a list, with first object contains p value, estimates, and CI of NIE1, NIE2, NIE, and each parameters.

A list contains all results.

`fullList`	A list contains resutls for each taxon. See details.
`validTaxaNames`	The taxons that were in group 1, which had enough zero and non-zero observations.
`continuTaxa`	The taxons that were in group 2, which only had non-zero observations.