classify.qtl {eqtl} | R Documentation |
Estimate wether an eQTL is cis- or trans- acting.
classify.qtl(cross, peak, etrait.coord, data.gmap)
cross |
An object of class cross . See 'qtl' package manual for read.cross function details. |
peak |
An object of class peak . See define.peak for details. |
etrait.coord |
A data.frame with column names "etrait.name","chr","start","stop" specifying the etrait (expression trait) location on the genome:etrait.coord$array_element_name is character strings vector specifying the name of the etraits.etrait.coord$chr is a vector of integers specifying the chromosome on which the markers are localized.etrait.coord$start is a vector of integers specifying the start of the etrait's sequence in base pair.etrait.coord$stop is a vector of integers specifying the stopo of the etrait's sequence in base pair. |
data.gmap |
A data.frame with column names "Marker" , "chr" and "PP" specifying the marker physical location. Those ones must be the same markers defined in the related cross object.data.gmap$Marker is a vector character strings specifying the names of markers.data.gmap$chr is a vector of integers specifying the chromosomes on which the markers are localized.data.gmap$PP is a vector of integers specifying the physical marker locations on the chromosomes in base pair. |
Useful in case of genome-wide expression QTL mapping. Determining cis-acting and trans-acting eQTL (or cis- and trans- eQTL) gives a basic overview about the global eQTL network. The (potential) cis-eQTL are those which colocalize with the controlled gene. These could be typically explained by a modification within gene promoter and therefore actually correspond to a cis-regulation (note that it would remain to be confirmed on a case by case basis: due to the lack of precision in QTLs localization for all analysis methods, a cis-acting is still biologically hypothetical; plus it could also correspond to a trans-acting eQTL localised close to its target gene). eQTLs which contains the regulated gene within their LOD support interval are classified in this category as cis
. The trans-acting eQTLs are defined as those which do not colocalize with the affected gene. These could typically correspond to the mode of action of a transcription factor on the regulation of another gene's expression. eQTL which do not contain the regulated gene within their LOD support interval are classified as trans
.
The input peak
object is returned with a component type
added to the components of names(peak$trait$chromosome)
for each previously detected QTL:
type |
cis or trans for cis- and trans- eQTL respectively.<NA> if the etrait location is unknown or not nuclear.
|
The QTL support interval locations are defined within a peak
object. This classification (performed by classify.qtl
) depends entirely on the support interval definition computed by the function define.peak
. This function tend to underestimate cis-eQTL number as LOD-drop value are more conservative. It does not replace the scientist's eye on the LOD curve.
Hamid A. Khalili
read.cross
,define.peak
,calc.adef
data(seed10); # Genotype probabilities seed10 <- calc.genoprob( cross=seed10, step=2, off.end=0, error.prob=0, map.function='kosambi', stepwidth='fixed'); seed10 <- sim.geno( cross=seed10, step=2, off.end=0, error.prob=0, map.function='kosambi', stepwidth='fixed'); # Genome scan and QTL detection out.em <- scanone( seed10, pheno.col=1:50, model='normal', method='em'); out.peak <- define.peak( out.em, 'all'); # Additive effect computing and peaks localization out.peak <- calc.adef(seed10,out.em,out.peak); data(BSpgmap); out.peak <- localize.qtl(seed10,out.peak,BSpgmap); # Estimated actind-type of the expression QTL affecting the 100th expression trait and # localized on chromosome 1 data(ATH.coord) out.peak <- classify.qtl(seed10,out.peak,ATH.coord,BSpgmap); out.peak[[26]]$'4'$type; # idem for the trait 'CATrck' out.peak$CATrck$'4'$type;