clac.preparenormal.R {clac} | R Documentation |
CLAC is a method for calling gains and losses in CGH array data. This function is to prepare normal references in CLAC(Cluster Along Chromosome) Analysis
clac.preparenormal.R(CANCER, NORMAL, Normal.Type, chromosome.number, nucleotide.position, windowsize=5, targetFDR=0.01, chromosomeOption=FALSE, centromere=NULL)
CANCER |
data matrix. It's the result of a group of CGH experiments, which are the target disease arrays to analysis. Each column corresponds to one sample (one target array), and each row corresponds to one gene/clone. The (i, j) entry should be the log fluorescence ratio of the ith gene/clone in the jth sample. Missing value should be coded as either NA or 999. |
NORMAL |
data matrix. It's also the result of a group of CGH experiment. But these results are from normal reference arrays. Again, each column corresponds to one sample, and each row corresponds to one gene/clone. Missing value should be coded as either NA or 999. |
Normal.Type |
a vector specifying the normal array type. Length should be the same as the column number of NORMAL . Code 0 for normal reference arrays from the same gender hybridization, while 1 for arrays from opposite gender hybridization. |
chromosome.number |
numeric vector . Length should be the same as the row number of NORMAL . It's the chromosome number of each gene/clone. |
nucleotide.position |
numeric vector. Length should be the same as the row number of NORMAL . It's the nucleotide position of each gene/clone. |
windowsize |
numeric value, specifying the window size to carry out the average smooth. |
targetFDR |
numeric value between 0 and 1, specifying the desired fianl FDR for CLAC analysis. |
chromosomeOption |
a boolean variable. If False, the chromosome arms will be considered seperately. If true, two chormosome arms of one chromosome would be dealed together. |
centromere |
numeric vector specifying the centromere positions. If missing, the default centromere value of human genome will be used. |
clac.preparenormal.R
builds cluster trees on normal reference arrays, the result is reported to the next step of CLAC analysis.
A list with components
normal.result |
An object containing the information about the normal reference arrays. |
CANCER.sm |
The result of average smooth for data matrix CANCER |
NORMAL.sm |
The result of average smooth for data matrix NORMAL |
Pei Wang
P. Wang, Y. Kim, J. Pollack, B. Narasimhan and R. Tibshirani, ¡°A method for calling gains and losses in array CGH data¡±, Biostatistics (accepted for publication 4/5/2004), available at http://www-stat.stanford.edu/~wp57/CGH-Miner/
library(clac) data(BACarray) attach(BACarray) ############ prepare the normal reference arrays NormalResult<-clac.preparenormal.R(DiseaseArray, NormalArray, Normal.Type=rep(0,3), chromosome.number=chromosome, nucleotide.position=nucposition, windowsize=5, targetFDR=0.01, chromosomeOption=FALSE) ############ clac on selected tumor arrays clac.result<-clac.tumorarray.R(NormalResult, tumorarrayIndex=1:4) ############ Plot for the first arrays i<-1 clac.PlotSingleArray.R(i, NormalResult, clac.result) title(main=paste("CLAC Plot for the ", i ,"th BAC array; FDR=", round(clac.result$fdr[i],3), sep="")) ############ consensus plot clac.PlotConsensus.R(clac.result, chromosome, nucposition, 1:4) title(main="Consensus Plot for 4 BAC arrays") ############ Plot all arrays clac.PlotAllArray.R(NormalResult, clac.result) title(main="Plot for all 4 arrays")