Zscore {metaArray} | R Documentation |
This function calculates Z-score for each matched gene across all datasets. In each dataset, it performs local regression smoothing of mean vs variance. Z score is constructed by taking the ratio of weighted mean difference and combined standard deviation according to Box and Tiao (1992).
Zscore(merged, pheno = NULL, permute = 0, verbose = TRUE)
merged |
mergeExprSet object that contains gene expression and class label with all
datasets. Class label should consist of two unique elements. If pheno is NULL, first columns
of phenoData from each ExpressionSet is sought as class labels. If a vector of particular column number
in each data is specified, corresponding columns of phenoData will be considered for class labels. |
pheno |
A numeric vector specifying the location of class labels in phenoData from each
ExpressionSet , a unit of mergeExprSet representing one dataset. |
permute |
If permute is 0, weighted Z-score will be referenced to standard normal distribution for two-sided p-value. Otherwise, columns of all datasets (each dataset separately) will be shuffled at random, from which a permutation distribution of Z-scores are formed and Z-scores are referenced to this distribution. |
verbose |
If verbose is TRUE, the progress of permutation will be reported. |
A data.frame with matched genes, Z-scores and p-values will result.
Debashis Ghosh <ghoshd@umich.edu>, Hyungwon Choi <hwchoi@umich.edu>
J.Wang et al, Bioinformatics 2004 Nov 22;20(17):3166-78
# Zscore(merged, pheno=NULL, permute=10000, verbose=FALSE)