bestscan: Make 1-D scans from best patterns

In byandell/qtlview: Visualization of QTL Scans for Multiple Traits

Description

Takes best model per trait and builds a matrix of 1-D scans.

Usage

make.bestscan(cross, best, pattern, scan.type = "BIM",
  cross.name = deparse(substitute(cross)),
  maps = get(paste(cross.name, "maps", sep = ".")),
  trait.annotation = get(paste(cross.name, "traitnames", sep = ".")),
  attenuate = 1, tolerance = 1e-05, verbose = TRUE)

Arguments

cross	Object created by read.cross.
best	Data frame with trait name as first column and chrom and locus as two other columns. Last column contains the QTL-specific score (e.g. explained variance). See details.
pattern	Data frame with trait name as first column, and columns for pattern and posterior.
scan.type	Type of scan as character string.
cross.name	Name of cross object as character string.
maps	Object of class read.maps.
trait.annotation	Annotation object for traits.
attenuate	Degree of attenuation of peaks.
tolerance	Tolerance level (drop scans below this level).
verbose	Show steps if TRUE.

Details

This converts results from an object that has the best pattern per trait to a matrix with 1-D scans, one trait per row. This is meant to be saved for use in multtrait, ideally filtered through a web page.

The best object has one row per QTL for each trait; traits with no QTL can be included with NA for chrom and locus. Other columns in best will depend on the method of construction, but the last column should contain a QTL-specific score, which is used for plotting. For instance, using qb.best from suggested package R/qtlbim includes a column for the posterior probability for that QTL (n.qtl), and a last column with the QTL-explained variance.

The pattern object is constructed as best as possible from best if not provided. The trait.annotation, created with read.annotation, is used for trait positions if provided (see aug.scanone).

The attenuate determines the degree of attenuation down from peak. Default is an exponential decay corresponding to genetic linkage in cM.

Author(s)

Brian S. Yandell

See Also

read.cross, read.maps, read.annotation

Examples

## Not run: 
tissue.bestscan <- make.bestscan(cross, tissue.best, tissue.pattern)

## End(Not run)

byandell/qtlview documentation built on May 13, 2019, 9:53 a.m.