eCCC_ext: The Extended e-CCC Algorithm

In ewouddt/CCCBiclust: R Wrapper for Java Implementation of CCC and e-CC biclustering algorithm

Description

A R-wrapper which directly calls the original Java code for the Extended e-CCC algorithm (http://kdbio.inesc-id.pt/software/e-ccc-biclustering/) and transforms it to the output format of the Biclust R package.

Usage

eCCC_ext(matrix, minr = 1, minc = 1, maxErrors = 1, overlap = 0.1,
  missings = "allow", anticorrelation = FALSE, restrictedErrors = FALSE)

Arguments

matrix	The input matrix in which the columns are ordered by time.
minr	Integer containing the row quorum (= minimum number of genes allowed in e-CCC-Biclusters).
minc	Integer containing the column quorum (= minimum number of contiguous time points allowed in e-CCC-Biclusters).
maxErrors	Integer containing the amount of errors allowed, per gene, in the e-CCC-Biclustering algorithm (value of e).
overlap	Numeric in ]0,1[ containing the maximum percentage of overlapping allowed. The info slot will contain a logical vector which can filter the biclusters which overlap more than overlap*100 percent.
missings	How to handle missing values: "remove": remove genes with missing values "allow": allow missing values as valid errors "jump": 'jump over' missing values
anticorrelation	Logical value to allow or not allow anticorrelation. If allowed (TRUE), the algorithm will look for e-CCC-Biclusters with Sign-Changes.
restrictedErrors	Logical value restricting or not restricting errors. If restricting (TRUE), errors are restricted to the symbols in the 1-neighbourhood of the symbols in the alphabet. Since the alphabet D,N,U is used in the predefined discretization step provided in this version of the prototype, the number of neighbours used in the restricted errors extension can only be equal to 1.

Details

PLACEHOLDER

Value

A Biclust S4 Class object containing extra information of the e-CCC algorithm result in the info slot.

Author(s)

Ewoud De Troyer

References

Sara C. Madeira and Arlindo L. Oliveira, "A polynomial time biclustering algorithm for finding genes with approximate expression patterns in gene expression time series", Algorithms for Molecular Biology 2009, 4:8 (4 June 2009)

Examples

## Not run: 
data(heatdata)
out <- eCCC_ext(heatdata,minr=3,minc=2,missings="jump",anticorrelation = TRUE)
eCCCinfo(out,filter="Bonf0.01")
out@info$eCCCpatterns$BC53[1:10,]

## End(Not run)

ewouddt/CCCBiclust documentation built on May 16, 2019, 9:41 a.m.