significantMicroRna: Summarize Differential Expression Results

In AgiMicroRna: Processing and Differential Expression Analysis of Agilent microRNA chips

Description

The function summarizes the results from the differential expression analysis using the different objects that are obtained after 'limma' has been used, such as the 'MArrayLM' object with the statistics and the 'TestResults' object highlighting the significant features.

Usage

significantMicroRna(eset, ddset, targets, fit2, 
CM, DE, DEmethod, MTestmethod, PVcut, Mcut,verbose=FALSE)

Arguments

eset	ExpressionSet containing the Total Gene processed data
ddset	An uRNAList object containing the Total Gene proceseed data
targets	data.frame with the target structure
fit2	MArrayLM object from eBayes 'limma' function
CM	Contrast matrix
DE	TestResults object
DEmethod	method used in decideTests, only 'separate' or 'nestedF' are implemented
MTestmethod	method for multiple test
PVcut	p value threshold to declare significant features
Mcut	M value threshold to select within significant features
verbose	logical, if TRUE prints out output

Details

A list containing the genes with their statistics is generated. The significant genes above the PVcut p values are also given in a html file that links the selected miRNAS to the miRBase http://microrna.sanger.ac.uk/. A MA plots indicating the differentially expressed genes are also displayed.

When multiple contrasts are done, the method for the selection of the significant genes can be either 'separated' or 'nestedF'. See decideTests in package limma limma for a detailed description on these two methods. When 'separated' is used a list with all the genes that have been analized in limma is given. The list includes de following columns:

PROBE - Probe name (one of the probes interrogating the gene) GENE - miRNA name PROBE chr\_coord - Agilent chromosomal location M - Fold change A - Mean of the intensity for that miRNA t - moderated t-statistic pval - p value of the t-statistic adj.pval - p value adjusted by 'MTestmethod' fdr.pval - p value adjusted by fdr

Some times, the user can be set 'MTestmethod = none', in this case, it might be interesting to still see the fdr value, despite of the fact that the user has decided not apply any multiple testing correction.

If the 'nestedF' is used, then two lists are provided for each contrasts. A first containing the selected significant genes, and a second list containing the rest of the genes that have been analyzed. The columns given in this case is:

PROBE - Probe name (one of the probes interrogating the gene) GENE - miRNA name PROBE chr\_coord - Agilent chromosomal location M - Fold change A - Mean of the intensity for that miRNA t - moderated t-statistic t pval - p value of the t-statistic F - F statistic (null hypothesis: Ci = Cj, for all contrasts i, j) adj.F.pval - F p value adjusted by 'MTestmethod' fdr.F.pval - F p value adjusted by fdr

The html files, both for the 'separated' and 'nestedF' method, includes only the selected as significant genes.

Author(s)

Pedro Lopez-Romero

References

Smyth, G. K. (2005). Limma: linear models for microarray data. In: 'Bioinformatics and Computational Biology Solutions using R and Bioconductor'. R. Gentleman, V. Carey, S. Dudoit, R. Irizarry, W. Huber (eds), Springer, New York, pages 397–420.

miRBase: the home of microRNA data http://microrna.sanger.ac.uk/

See Also

A 'uRNAList' example containing proccesed data is in ddPROC and an overview of how the processed data is produced is given in filterMicroRna. The ExpressionSet object can be generated using esetMicroRna An overview of miRNA differential expression analysis is given in basicLimma An example of how to get the 'TestResults' object is in getDecideTests

Examples

data(targets.micro)
data(ddPROC)
esetPROC=esetMicroRna(ddPROC,targets.micro,makePLOT=FALSE)

levels.treatment=levels(factor(targets.micro$Treatment))
treatment=factor(as.character(targets.micro$Treatment),
    levels=levels.treatment)
            
levels.subject=levels(factor(targets.micro$Subject))
subject=factor(as.character(targets.micro$Subject),
    levels=levels.subject)
	
design=model.matrix(~ -1 + treatment + subject  )

CM=cbind(MSC_AvsMSC_B=c(1,-1,0,0),
          MSC_AvsMSC_C=c(1,0,-1,0))
                
fit2=basicLimma(esetPROC,design,CM,verbose=TRUE)

DE=getDecideTests(fit2,
        DEmethod="separate",
        MTestmethod="BH",
        PVcut=0.10)

significantMicroRna(esetPROC,
	ddPROC,
	targets.micro,
  	fit2,
	CM,
	DE,
	DEmethod="separate",
	MTestmethod="BH",
	PVcut=0.10,
	Mcut=0,
	verbose=TRUE)

AgiMicroRna documentation built on Nov. 8, 2020, 5:25 p.m.