A QTL cis/trans classifier

Usage

qtlClassifier(geneSNP, SNPPos, genePos, radius)

Arguments

geneSNP: data.frame with columns on gene, SNP and biomarker (e.g., expression, protein).
SNPPos: data.frame containing SNP, chromosome and position.
genePos: data.frame containing gene, chromosome, start and end positions.
radius: flanking distance.

Value

It returns a geneSNP-prefixed data.frame with the following columns:

geneChrom gene chromosome.
geneStart gene start.
geneEnd gene end.
SNPChrom pQTL chromosome.
SNPPos pQTL position.
Type cis/trans labels.

Details

The function obtains QTL (simply called SNP here) cis/trans classification based on gene positions.

Note

This is adapted from iBMQ/eqtlClassifier as an xQTL (x=e, p, me, ...) classifier.

Examples

if (FALSE) { # \dontrun{
  merged <- read.delim("INF1.merge",as.is=TRUE)
  hits <- merge(merged[c("CHR","POS","MarkerName","prot","log10p")],
                inf1[c("prot","uniprot")],by="prot")
  names(hits) <- c("prot","Chr","bp","SNP","log10p","uniprot")

  options(width=200)
  geneSNP <- merge(hits[c("prot","SNP","log10p")],
                   inf1[c("prot","gene")],by="prot")[c("gene","SNP","prot","log10p")]
  SNPPos <- hits[c("SNP","Chr","bp")]
  genePos <- inf1[c("gene","chr","start","end")]
  cvt <- qtlClassifier(geneSNP,SNPPos,genePos,1e6)
  cvt
  cistrans <- cis.vs.trans.classification(hits,inf1,"uniprot")
  cis.vs.trans <- with(cistrans,data)
  cistrans.check <- merge(cvt[c("gene","SNP","Type")],cis.vs.trans[c("p.gene","SNP","cis.trans")],
                          by.x=c("gene","SNP"),by.y=c("p.gene","SNP"))
  with(cistrans.check,table(Type,cis.trans))
} # }