DiffAnalysis: Differential Analysis for paired data

In anapuce: Tools for Microarray Data Analysis

Description

Performs the differential analysis from normalised paired microarray data according to different ways of variance modelling and computes lists of differentially expressed genes according different multiple testing procedures.

Usage

DiffAnalysis(fileIN = "resNorm.txt", n = 3, ind.array = 1:2, name.A = "A",
name.M = "M.norm", fileOUT = "ListOfGenes.txt",
fileDelete = "GenesOutOfAnalysis.txt", procs = c("bonferroni", "BH"),
alpha = c(0.05, 0.05), dyeswap = TRUE, indDS = c(2), fileID = NULL,
function.trt = NULL, by.var = "ID", varmixt.meth = TRUE, header = TRUE,
sep = "\t", sep.write = "\t", dec.write = ".", ...)

Arguments

fileIN	normalized data set.
n	number of identificant columns.
ind.array	the indices of arrays to analyze.
name.A	character string containing a regular expression contained in the columnames corresponding to the A values.
name.M	character string containing a regular expression contained in the columnames corresponding to the M values.
fileOUT	output data file.
fileDelete	output data file containing the list of withdrawn genes.
procs	adjusting p-values procedures to be used.
alpha	values of the first type error to be used in the different procedures.
dyeswap	logical indicating whether the design is a dye swap.
indDS	index of arrays related to dye swap.
fileID	file giving information about genes.
function.trt	function to be applied before differential analysis.
by.var	argument passed to function.trt.
varmixt.meth	logical indicating whether to perform the variance modelling proposed by Delmar et al. (2005).
header	a logical value indicating whether the file contains the names of the variables as its first line. Used in the read.table function.
sep	the field separator string to use in the read.table function.
sep.write	the field separator string to use in the write.table function.
dec.write	the string to use for decimal points in the write.table function.
...	Further arguments to be passed to read.table.

Details

This function performs a differential analysis in the gene-specific and homoscedastic cases. If varmixt.meth = TRUE, the method used is the method proposed by Delmar et al. (2005), (the same as coded in the vm.analysis.paired function from the varmixt package.)

Value

By default these following files are created

- one list of differentially expressed genes by procedure

- The list of withdrawn genes

- The list of all the genes

An R object is returned if function return value is assigned.

Author(s)

J. Aubert

References

Delmar, P., Robin, S. and Daudin, J.J., (2005), VarMixt: efficient variance modelling for the differential analysis of replicated gene expression data, Bioinformatics, 21,(4), 502–8

Dudoit, S., Yang, Y. H., Callow, M. J. and Speed, T.P., (2002), Statistical methods for identifying differentially expressed genes in replicated cdna microarray experiments, Statisticia Sinica, 12, 111–139

See Also

p.adjust, MeanBySpot, DiffAnalysis.unpaired

Examples

data(spleen)
# Analysis on the first 100 genes
resSpleen <- DiffAnalysis(spleen[1:100, ], n = 1, ind.array = 1:6, name.A = "A.", name.M = "M.",
varmixt.meth = TRUE, dyeswap = TRUE, indDS = c(2,4,6))
# Histogramm of the raw-pvalues
# hist(resSpleen$PValueVM)

Example output

####################################### 
DIFFERENTIAL ANALYSIS 
Arrays used in the analysis :  1 2 3 4 5 6 
Outfile for lists of genes :  ListOfGenes.txt 
Outfile for genes not analysed :  GenesOutOfAnalysis.txt 
Multiple testing procedures :  bonferroni BH 
Nominal Type I error rate for each procedure :  0.05 0.05 
####################################### 
Number of genes with a null variance :  0 
##########################################################
                  One variance by gene                    
##########################################################
Multiple testing procedure :  bonferroni (5%)  
Number of differentially expressed genes :  0 
Multiple testing procedure :  BH (5%)  
Number of differentially expressed genes :  0 
##########################################################
       A common variance for all the genes                
##########################################################
Number of dropped genes :  4 
Number of dropped genes with a too small variance :  0 
Number of dropped genes with a too high variance :  4 
 Estimated variance (standard error):  0.02630932 ( 0.03406337 ) 
Multiple testing procedure :  bonferroni (5%)  
Number of differentially expressed genes :  4 
Multiple testing procedure :  BH (5%)  
Number of differentially expressed genes :  7 
############################################################
 Clusters of genes with equal variance (Delmar et al. 2005)
############################################################
N mixt=1. EM algo, number of iterations=1. Elapsed time=0 minute(s).
pi= 1 
var= 0.04998202 
log-likelihood= -57.54241 
BIC= 119.69 

N mixt=2. EM algo, number of iterations=9. Elapsed time=0 minute(s).
pi= 0.8595523 0.1404477 
var= 0.01717431 0.2507681 
log-likelihood= 80.8037 
BIC= -147.7919 

N mixt=3. EM algo, number of iterations=45. Elapsed time=0 minute(s).
pi= 0.7550745 0.1903421 0.05458344 
var= 0.0145965 0.06044781 0.502988 
log-likelihood= 89.97303 
BIC= -156.9202 

N mixt=4. EM algo, number of iterations=33. Elapsed time=0 minute(s).
pi= 0.2619263 0.4941419 0.1893917 0.05454012 
var= 0.01388069 0.01501027 0.06063094 0.5032278 
log-likelihood= 89.97342 
BIC= -147.7106 

N mixt=3. EM algo, number of iterations=17. Elapsed time=0 minute(s).
pi= 0.7613167 0.1845434 0.05413996 
var= 0.01470256 0.06188923 0.5054949 
log-likelihood= 89.97812 
BIC= -156.9304 
----------------------------------------- 
Group  1  
   Number of genes  81 
   Variance         0.01470256 
Group  2  
   Number of genes  14 
   Variance         0.06188923 
Group  3  
   Number of genes  5 
   Variance         0.5054949 
----------------------------------------- 
Multiple testing procedure :  bonferroni (5%)  
Number of differentially expressed genes - VM :  1 
-------------------------------------------------------------
Multiple testing procedure :  BH (5%)  
Number of differentially expressed genes - VM :  8 
-------------------------------------------------------------

anapuce documentation built on May 2, 2019, 9:44 a.m.