calcCis: Calculate all confidence intervals from a QTL model

Description Usage Arguments Details Value Examples

Description

calcCis Takes a multiple QTL model or scanone and permutations output and returns a dataframe of confidence intervals.

Usage

1
2
3
calcCis(cross, mod = NULL, s1.output = NULL, pheno.col = NA,
  perm.output = NULL, qtlnames = NULL, lodint = TRUE, drop = 1.5,
  prob = 0.95, expandtomarkers = FALSE, chr = NULL, ...)

Arguments

cross

The QTL cross object.

mod

A QTL model containing LOD profiles. Can be generated by stepwiseqtl (with keepLodProfile = TRUE), or by refineqtl.

s1.output

Output from a scanone call. Not needed if qtl models are supplied to mod. If mod is not supplied, it is necessary to provide both s1.output and perm.output.

pheno.col

If passing a QTL model, specify the name of the phenotype.

perm.output

Output from a scanone call with n.perm > 1. Needs to have the identical parameters as s1.output.

qtlnames

An optional vector of user-specified QTL identifiers. Used only if mod is supplied.

lodint

Should lodint (default) or bayesint be used to estimate confidence intervals?

drop

If lodint = TRUE, what drop threshold should be used?

prob

If lodint = FALSE, what significance should be used for bayesint?

expandtomarkers

Logical. Should the intervals be expanded to the nearest real marker?

...

Additional arguments passed on to pullSigQTL, such as the alpha threshold for a qtl.

Details

This function iterates through the qtl in the model, returning confidence intervals.

Value

A dataframe of confidence intervals, including the maximum LOD score, bounding markers and positions.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
library(qtlTools)
data(fake.bc)
cross<-fake.bc
cross <- calc.genoprob(cross, step=2.5)

# Make a QTL object and formula
mod <- makeqtl(cross, chr = c(2,5), pos = c(40,25), what = "prob")
sex <- data.frame(sex = cross$phe$sex)
nform <- "y ~ Q1 + Q2 + Q1*sex + sex"

#Calculate lodprofiles for confidence interval estimation
mod <- refineqtl(cross, mod, pheno.col = "pheno1",
  qtl = mod, formula = nform, covar = sex, method="hk")
calcCis(cross = cross, mod=mod)
## Not run: 
s1<-scanone(cross, method="hk", pheno.col=c("pheno1", "pheno2"))
perm<-scanone(cross, n.perm=100, method="hk",pheno.col=c("pheno1", "pheno2"), verbose=FALSE)
calcCis(cross,s1.output=s1, perm.output=perm)

## End(Not run)

jtlovell/qtlTools documentation built on May 20, 2019, 3:14 a.m.