Learn R Programming

MPINet (version 1.0)

getPSS: Evaluate the CGNB score of metabolites

Description

Integrate the non-equivalence scores and the initial bias scores of metabolites by the monotonic spline model.

Usage

getPSS(riskmeta,plot=TRUE,binsize=400)

Arguments

riskmeta
A character vector of interesting metabolites, for each element is a pubchem CID.
plot
A logical. If TRUE the fit line obtained by the monotonic spline model will be plotted.
binsize
plot the fraction of differential metabolites in bins of this size.

Value

A data frame with 4 columns including "riskmeta", "meanstrvalue", "pss" and "CGNB". Each row correspond a metabolite pubchem CID. "riskmeta" indicates whether the metabolite is in the interesting set (with "1" is in and "0" is not in)."meanstrvalue" is the mean SOC value of the metabolite. "pss" is the score value obtained by the monotonic spline model. "CGNB" is the CGNB score of metabolite which is calculated as 1 substract the score value obtained by monotonic spline model. This score is used to calculate pathway weight in the subsequent pathway analysis.

Details

This function is used to integrate the non-equivalence of metabolites and the character of differential metabolites.The binsize can be assigned according to the best visualization.

References

Young, M.D., Wakefield, M.J., Smyth, G.K. and Oshlack, A. (2010) Gene ontology analysis for RNA-seq: accounting for selection bias. Genome Biol, 11, R14.

Examples

Run this code
## Not run: 
# 
# #######################################
# 
# 
# ##########get example data
# risk<-GetExampleData(dataset="prostate")
# 
# ###########calculate the CGNB score
# pss<-getPSS(risk ,plot=F)
# CGNBscore<-pss[,"CGNB"]
# names(CGNBscore)<-rownames(pss)
# ##########print the CGNB score of some metabolites to screen
# head(CGNBscore)
# #identify dysregulated pathways
# anncpdpre<-identifypathway(risk,pss,pathType="KEGG",method="MPINet",annlim=1,bglim=6)
# #convert ann to data.frame
# result<-printGraph(anncpdpre,pathType="KEGG",method="MPINet")
# head(result)
# 
# 
# 
# 
# 
# 
# ## End(Not run)

Run the code above in your browser using DataLab