plotP
SPIA two-way evidence plot
Plots each pathway as a point, using the over-representation p-value, pNDE, and perturbations accumulation p-value, pPERT, as coordinates. In addition the regions where FDR and FWER adjusted pG values are less than the specified threshold are plotted. The function determines automatically which method (fisher or norminv) was used to combine the two p-values into pG, and plots the regions described above accordingly.
- Keywords
- methods, nonparametric
Usage
plotP(x,threshold=0.05)
Arguments
- x
- A data frame produced by
spia
function. - threshold
- A numerical value between 0 and 1 to be used as significance threshold in inferring pathway significance.
Details
In this plot each pathway is a point and the coordinates are the log of pNDE (using a hypergeometric model) and the p-value from perturbations, pPERT. The oblique lines in the plot show the significance regions based on the combined evidence.
Value
-
This function does not return any value. It only generates a plot.
References
Adi L. Tarca, Sorin Draghici, Purvesh Khatri, et. al, A Signaling Pathway Impact Analysis for Microarray Experiments, 2008, Bioinformatics, 2009, 25(1):75-82.
See Also
Examples
# Examples use colorectal cancer dataset
data(colorectalcancer)
# pathway analysis based on combined evidence of ORA and perturbations
# use nB=2000 or larger for more accurate results
res<-spia(de=DE_Colorectal, all=ALL_Colorectal, organism="hsa",nB=200,plots=FALSE,verbose=TRUE,beta=NULL,combine="fisher")
#Generate the evidence plot
plotP(res,threshold=0.1)
res<-spia(de=DE_Colorectal, all=ALL_Colorectal, organism="hsa",nB=200,plots=FALSE,verbose=TRUE,beta=NULL,combine="norminv")
#Generate the evidence plot
plotP(res,threshold=0.1)