QTLres: Function to assess the performance of the eQTLs that BADGER...

In BADGER: A package for identifying sample mixups in combined expression and genotyping studies.

Description

This function compares the predicted and observed genotypes within BADGER with a view to informing refinement of the eQTL list

Usage

QTLres(predictedMATSNP, MATSNP)

Arguments

predictedMATSNP	A numeric matrix of predicted genotypes (probably obtained from BADGERpredSNP
MATSNP	A matching numeric matrix of observed genotypes.

Details

Based on Euclidean distance.

Value

Numeric vector, each entry giving a measure of the predictive ability of an eQTL

Author(s)

Lynch AG

References

First used/described in

Lynch et al. "Ensuring data integrity in large modern integrative genomic studies" ECCB 2010, Ghent, Poster abstract. http://www.ebi.ac.uk/eccb/2010/index.php/proceedings/posters/122-other-bioinformatics-applications.html

Curtis et al. "The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups" Nature 486, 346-352

Lynch et al. "Calling sample mixups in cancer population studies" submitted

http://badger.r-forge.r-project.org/

See Also

BADGER,BADGERmatchScores,BADGERpredSNP

Examples

data(exampleEXP)
data(exampleSNP)
mysampnames<-sapply(colnames(exampleEXP),function(x){unlist(strsplit(as.character(x),"_"))[2]})
mypred<-BADGERpredSNP(exampleEXP[,match(colnames(exampleSNP),mysampnames)],exampleSNP,exampleEXP)
myQTLres<-QTLres(mypred[,match(colnames(exampleSNP),mysampnames)],exampleSNP)
myMS<-BADGERmatchScores(mypred,exampleSNP,which(myQTLres<35))

mycheck<-rep(0,422) 
for(i in 1:422){mycheck[i]<-colnames(exampleSNP)[which.min(myMS[i,])]} 
mycheck[!(mysampnames %in% colnames(exampleSNP))]<-NA

# see that there is a problem with the expression matrices associated with NA18515

cbind(mycheck,mysampnames)[which(mycheck!=mysampnames),]

BADGER documentation built on May 2, 2019, 4:56 p.m.