The BCDiag package

Description

Bicluster Diagnostics plots

Introduction

The Bicluster Diagnostics plots(BcDiag) package is a visualization technique, for profiling and summarizing Bicluster data, particularly for gene expression level data. Target data matrix are bicluster genes(rows) and conditions(columns) versus clustered genes or conditions.

Main task

A BicDiag is a package of visualaization bicluster data, which is a subset matrix that have similar characterstics in terms of row(genes) and columns(conditions).

It has used three different types of bicluster algorithms to extract the biculsterd data; 'biclust','isa2' and 'fabia'. plots such as boxplot,histogram, line plot,3D plot are some of the plots that have used to visualize the data.

Major taskes of the package can be categorized in to three sections;

1. profiling and summarizing the biclustered vs. the clustered simultaneously

2. profiling and summarizing only the biclusterd data.

3. exploring the biclusterd data using anova and median polish techniques.

Author(s)

Mengsteab Aregay mycs.zab@gmail.com

References

Hochreiter, S., Bodenhofer, U., Heusel, M.et al. (2010).FABIA: factor analysis for bicluster acquisition. Bioinformatices, 26, 1520-1527.

Kaiser S. and Leisch F. (2008). A Toolbox for Bicluster Analysis in R. Ludwigstrasse. 33.

Csardi G., Kutalik Z., and Bergmann S.(2010). Modular analysis of gene expression data with R. Bioinformatics, 26, 1376-7

See Also

The Bicluster algorithms in the packages biclust,fabia and isa2.

The anomedOnlybic function

Description

Provides ANOVA and median polish residual plots for biclustered data.

Usage

anomedOnlybic(dset, bres, fit="boxplot", mname="biclust", bnum=1, 
fabia.thresZ=0.5,fabia.thresL=NULL)

Arguments

Details

A function provides residuals plots for biclustered data based on ANOVA and median polish.

The function checks the required parameter values and fit the plot according to the user requirements.

Note that the "biclust" option for mname will also accept results from the packages iBBiG and rqubic.

Value

Residual plots or residual box plots.

Author(s)

Mengsteab Aregay mycs.zab@gmail.com

References

Van't Veer, L.J., Dai, H., van de Vijver, M.J., He, Y.D., Hart, A.A. et al. (2002). Gene expression profiling predicts clinical outcome of breast cancer,Nature, 415, 530-536.

Kaiser S. and Leisch F. (2008). A Toolbox for Bicluster Analysis in R. Ludwigstrasse. 33.

Examples

data(breastc)
library(biclust)
# find bicluster using one of biclust algorithms

bic <- biclust(breastc, method=BCPlaid())
# fit residual boxplot from ANOVA
anomedOnlybic(dset=breastc,bres=bic,fit="boxplot",mname="biclust")

Gene Expression Data Example

Description

Microarray data set of van't Veer breast cancer.

Usage

data(breastc)

Format

A data matrix with 1213 genes and 97 samples.

References

Van't Veer, L.J., Dai, H., van de Vijver, M.J., He, Y.D., Hart, A.A. et al. (2002). Gene expression profiling predicts clinical outcome of breast cancer,Nature, 415, 530-536.

Examples

  
data(breastc)
  
head(breastc)

Gene Expression Data Example

Description

Log transformed Microarray data set of Rosenwald diffuse large-B-cell lymphoma.

Usage

data(dlbcl)

Format

A data matrix with 661 genes and 141 samples.

References

Rosenwald, A., Wright, G., Chan, W.C., Connors, J.M., Campo, E. et al. (2002). The use of molecular profiling to predict survival after chemotherapy for diffuse large-B-cell lymphoma, New Engl. J. Med., 346. 1937-1947.

Examples

  
data(dlbcl)
  
head(dlbcl)

The exploreBic function

Description

Provides exploratory plots for biclustered and clustered data.

Usage

exploreBic(dset, bres, gby ="genes", pfor ="mean", mname ="biclust", bnum =1, 
fabia.thresZ=0.5,fabia.thresL=NULL)

Arguments

Details

The exploreBic function is mainly used for exploratory data analysis. It provides summary plots for mean, median, variance, MAD and quantile plot.

The exploreBic function checks if the parameters are appropriately submitted and then identifies the biclusters submatrix and calculates its summary statistics. Finally, the results are displayed on the required plot.

Note that the "biclust" option for mname will also accept results from the packages iBBiG and rqubic.

Value

Summary plot will display according to the user specification.

Author(s)

Mengsteab Aregay mycs.zab@gmail.com

References

Van't Veer, L.J., Dai, H., van de Vijver, M.J., He, Y.D., Hart, A.A. et al. (2002). Gene expression profiling predicts clinical outcome of breast cancer,Nature, 415, 530-536.

Hochreiter, S., Bodenhofer, U., Heusel, M.et al. (2010).FABIA: factor analysis for bicluster acquisition. Bioinformatices, 26, 1520-1527.

See Also

exploreOnlybic

Examples

data(breastc)
# find bicluster using biclust package
library(biclust)
bic <- biclust(breastc,method=BCPlaid())
# Plot the mean of biclusterd and clustered genes parallely.
exploreBic(dset=breastc,bres=bic,gby="conditions",pfor="mean",mname="biclust")

The exploreOnlybic function

Description

Provides exploratory plots only for biclustering results.

Usage

exploreOnlybic(dset, bres, pfor= "all", gby= "genes", mname="biclust",bnum=1, 
fabia.thresZ=0.5,fabia.thresL=NULL)

Arguments

Details

The exploreOnlybic function has similar function with exploreBic. The only difference is that it provides exploratory plots only for biclustered data.

Value

Summary plot will display only for biclustered data.

Note that the "biclust" option for mname will also accept results from the packages iBBiG and rqubic.

Author(s)

Mengsteab Aregay mycs.zab@gmail.com

References

Van't Veer, L.J., Dai, H., van de Vijver, M.J., He, Y.D., Hart, A.A. et al. (2002). Gene expression profiling predicts clinical outcome of breast cancer,Nature, 415, 530-536.

Hochreiter, S., Bodenhofer, U., Heusel, M.et al. (2010).FABIA: factor analysis for bicluster acquisition. Bioinformatices, 26, 1520-1527.

See Also

exploreBic

Examples

data(breastc)
# find bicluster using biclust algorithm
library(biclust)
bic <- biclust(breastc,method=BCPlaid())
# Plot the median of biclusterd data.
exploreOnlybic(dset=breastc, bres=bic, pfor="all", gby="genes", mname="biclust", bnum=1)

The profileBic function.

Description

Provides profile plots for biclustered and clustered data.

Usage

profileBic(dset, bres, mname = c("fabia", "isa2", "biclust","bicare"), bplot = "all",
gby = "genes", bnum = 1, teta = 120, ph = 30, fabia.thresZ=0.5,fabia.thresL=NULL,
BClabel=TRUE,gene.lines=NULL,condition.lines=NULL)

Arguments

Details

The profile.bic function checks if all parameters are correctly submitted and then identifies the biclustered and clustered data.

Note that the "biclust" option for mname will also accept results from the packages iBBiG and rqubic.

Value

profile.bic(dset, bres, mname="biclust", bplot="all", gby="genes", bnum=1, teta=120, ph=30)

Author(s)

Mengsteab Aregay mycs.zab@gmail.com

References

Van't Veer, L.J., Dai, H., van de Vijver, M.J., He, Y.D., Hart, A.A. et al. (2002). Gene expression profiling predicts clinical outcome of breast cancer,Nature, 415, 530-536.

Kaiser S. and Leisch F. (2008). A Toolbox for Bicluster Analysis in R. Ludwigstrasse. 33.

Examples

# create the biclustering result
data(breastc)
library(biclust)
bic<- biclust(breastc, method=BCPlaid())
# 3 biclusters found

# 3D profile plot for biclustered and clustered data.
profileBic(dset=breastc,bres=bic,mname="biclust",
bplot="3D",gby="genes",teta=-30,ph=50,bnum=1)

The writeBic function

Description

Provides a summary output in a text format, extracted from 'biclust','isa2' and 'fabia' bicluster algorithms.

Usage

writeBic(dset, fileName, bicResult, bicname, 
mname = c("fabia", "isa2", "biclust","bicare"), append = TRUE, delimiter = " ", 
fabia.thresZ=0.5,fabia.thresL=NULL)

Arguments

Details

The original function was developed in 'biclust' package by Kaiser et.al (2008). We extend the function to be used for further bicluster algorithms, such as; 'isa2', 'fabia' and 'bicare'.

Note that the "biclust" option for mname will also accept results from the packages iBBiG and rqubic.

Value

Biclustered text file with title, total number of biclustered, dimension and name of the biclustered genes(rows) or conditions(columns).

Author(s)

Mengsteab Aregay

mycs.zab@gmail.com

References

Van't Veer, L.J., Dai, H., van de Vijver, M.J., He, Y.D., Hart, A.A. et al. (2002). Gene expression profiling predicts clinical outcome of breast cancer,Nature, 415, 530-536.

Kaiser S. and Leisch F. (2008). A Toolbox for Bicluster Analysis in R. Ludwigstrasse. 33.

Csardi G., Kutalik Z., and Bergmann S.(2010). Modular analysis of gene expression data with R. Bioinformatics, 26, 1376-7

See Also

biclust

Examples

# create the biclustering result
data(breastc)
library(fabia)
fab<- fabia(breastc)
# write the biclustering result into a text file
writeBic(dset=breastc,fileName="fabiabreast.txt",
bicResult=fab, bicname="Biclust results for fabia",
mname="fabia")