qtl (version 1.70)

scanPhyloQTL: Single-QTL genome scan to map QTL to a phylogenetic tree

Description

Jointly consider multiple intercrosses with a single diallelic QTL model, considering all possible partitions of the strains into the two QTL allele groups.

Usage

scanPhyloQTL(crosses, partitions, chr, pheno.col=1,
             model=c("normal", "binary"), method=c("em", "imp", "hk"),
             addcovar, maxit=4000, tol=0.0001, useAllCrosses=TRUE,
             verbose=FALSE)

Value

A data frame, as for the output of scanone, though with LOD score columns for each partition that is considered. The result is given class "scanPhyloQTL".

Arguments

crosses

A list with each component being an intercross, as an object of class cross (see read.cross for details). The names (of the form "AB") indicate the strains in the cross.

partitions

A vector of character strings of the form "AB|CD" or "A|BCD" indicating the set of paritions of the strains into two allele groups. If missing, all partitions should be considered.

chr

Optional vector indicating the chromosomes for which LOD scores should be calculated. This should be a vector of character strings referring to chromosomes by name; numeric values are converted to strings. Refer to chromosomes with a preceding - to have all chromosomes but those considered. A logical (TRUE/FALSE) vector may also be used.

pheno.col

Column number in the phenotype matrix which should be used as the phenotype. This can be a vector of integers; for methods "hk" and "imp" this can be considerably faster than doing them one at a time. One may also give a character strings matching the phenotype names. Finally, one may give a numeric vector of phenotypes, in which case it must have the length equal to the number of individuals in the cross, and there must be either non-integers or values < 1 or > no. phenotypes; this last case may be useful for studying transformations.

model

The phenotype model: the usual normal model or a model for binary traits

method

Indicates whether to use the EM algorithm, imputation, or Haley-Knott regression.

addcovar

Optional set of additive covariates to include in the analysis, as a list with the same length as crosses. They must be numeric vectors or matrices, as for scanone.

maxit

Maximum number of iterations for method "em".

tol

Tolerance value for determining convergence for method "em".

useAllCrosses

If TRUE, use all crosses in the analysis of all partitions, with crosses not segregating the QTL included in the estimation of the residual variance.

verbose

If TRUE, print information about progress.

Author

Karl W Broman,

Details

The aim is to jointly consider multiple intercrosses to not just map QTL but to also, under the assumption of a single diallelic QTL, identify the set of strains with each QTL allele.

For each partition (of the strains into two groups) that is under consideration, we pull out the set of crosses that are segregating the QTL, re-code the alleles, and combine the crosses into one large cross. Crosses not segregating the QTL are also used, though with no QTL effects.

Additive covariate indicators for the crosses are included in the analysis, to allow for the possibility that there are overall shifts in the phenotypes between crosses.

References

Broman, K. W., Kim, S., An\'e, C. and Payseur, B. A. Mapping quantitative trait loci to a phylogenetic tree. In preparation.

See Also

plot.scanPhyloQTL, summary.scanPhyloQTL, max.scanPhyloQTL, inferredpartitions, simPhyloQTL

Examples

set.seed(66787403)
# example map; drop X chromosome
data(map10)
map10 <- map10[1:19]
map10 <- map10[1:4]

# simulate data
x <- simPhyloQTL(4, partition="AB|CD", crosses=c("AB", "AC", "AD"),
                 map=map10, n.ind=150,
                 model=c(1, 50, 0.5, 0))

# run calc.genoprob on each cross
if (FALSE) x <- lapply(x, calc.genoprob, step=2)
x <- lapply(x, calc.genoprob, step=0)

# scan genome, at each position trying all possible partitions
out <- scanPhyloQTL(x, method="hk")

# maximum peak
max(out, format="lod")

# approximate posterior probabilities at peak
max(out, format="postprob")

# all peaks above a threshold for LOD(best) - LOD(2nd best)
summary(out, threshold=1, format="lod")

# all peaks above a threshold for LOD(best), showing approx post'r prob
summary(out, format="postprob", threshold=3)

# plot results
plot(out)