BMIQ: BMIQ
In RnBeads: RnBeads

Description Usage Arguments Details Value Author(s)

This function makes 3 indipendent attempts to fit a 3-state beta mixture model on the provided type I probes. An attempt is successful if at least 4 probes are assigned to each level. In case all attempts fail, the return value is NULL.

BMIQ(
  beta.v,
  design.v,
  doH = TRUE,
  nfit = 50000,
  th1.v = c(0.2, 0.75),
  th2.v = NULL,
  niter = 5,
  tol = 0.001
)

`beta.v`	`double` vector consisting of beta values. Missing values (`NA`s) cannot be handled, so these must be removed or imputed prior to running BMIQ. Before normalization, beta values that are exactly 0 and exactly 1 are replaced by the minimum positive and maximum value below 1, respectively.
`design.v`	`integer` vector of length `length(beta.v)`, containing the values `1` and `2` only. These values specify probe design type.
`doH`	Flag indicating if normalization for hemimethylated type II probes is to be performed.
`nfit`	Number of probes of a given design to use for the fitting. Smaller values will make BMIQ faster at the expense of accuracy. Values between 10000 and 50000 seem to work well.
`th1.v`	Thresholds "type 1" to use for the initialization of the EM algorithm. These values should represent best guesses for calling type I probes hemi-methylated and methylated, and are refined in further steps by the algorithm.
`th2.v`	Thresholds "type 2" to used for the initialization of the EM algorithm. These values should represent best guesses for calling type II probes hemi-methylated and methylated, and are refined in further steps by the EM algorithm. If this is `NULL` (default), the thresholds are estimated based on `th1.v` and a modified PBC correction method.
`niter`	Maximum number of EM iterations to be performed.
`tol`	Tolerance threshold for EM algorithm.