Package 'qtl'

Description

A brief introduction to the R/qtl package, with a walk-through of an analysis.

New to R and/or R/qtl?

• In order to use the R/qtl package, you must type (within R) library(qtl). You may wish to include this in a .Rprofile file.

• Documention and several tutorials are available at the R archive (https://cran.r-project.org).

• Use the help.start function to start the html version of the R help.

• Type library(help=qtl) to get a list of the functions in R/qtl.

• Use the example function to run examples of the various functions in R/qtl.

• A tutorial on the use of R/qtl is distributed with the package and is also available at https://rqtl.org/rqtltour.pdf.

• Download the latest version of R/qtl from the R archive or from https://rqtl.org.

Walk-through of an analysis

Here we briefly describe the use of R/qtl to analyze an experimental cross. A more extensive tutorial on its use is distributed with the package and is also available at https://rqtl.org/rqtltour.pdf.

A difficult first step in the use of most data analysis software is the import of data. With R/qtl, one may import data in several different formats by use of the function read.cross. The internal data structure used by R/qtl is rather complicated, and is described in the help file for read.cross. We won't discuss data import any further here, except to say that the comma-delimited format ("csv") is recommended. If you have trouble importing data, send an email to Karl Broman, [email protected], perhaps attaching examples of your data files. (Such data will be kept confidential.) Also see the sample data files and code at https://rqtl.org/sampledata/.

We consider the example data hyper, an experiment on hypertension in the mouse, kindly provided by Bev Paigen and Gary Churchill. Use the data function to load the data.

data(hyper)

The hyper data set has class "cross". The function summary.cross gives summary information on the data, and checks the data for internal consistency. A number of other utility functions are available; hopefully these are self-explanatory.

summary(hyper)
nind(hyper)
nphe(hyper)
nchr(hyper)
nmar(hyper)
totmar(hyper)

The function plot.cross gives a graphical summary of the data; it calls plotMissing (to plot a matrix displaying missing genotypes) and plotMap (to plot the genetic maps), and also displays histograms or barplots of the phenotypes. The plotMissing function can plot individuals ordered by their phenotypes; you can see that for most markers, only individuals with extreme phenotypes were genotyped.

plot(hyper)
plotMissing(hyper)
plotMissing(hyper, reorder=TRUE)
plotMap(hyper)

Note that one marker (on chromosome 14) has no genotype data. The function drop.nullmarkers removes such markers from the data.

hyper <- drop.nullmarkers(hyper)
totmar(hyper)

The function est.rf estimates the recombination fraction between each pair of markers, and calculates a LOD score for the test of $r$ = 1/2. This is useful for identifying markers that are placed on the wrong chromosome. Note that since, for these data, many markers were typed only on recombinant individuals, the pairwise recombination fractions show rather odd patterns.

hyper <- est.rf(hyper)
plotRF(hyper)
plotRF(hyper, chr=c(1,4))

To re-estimate the genetic map for an experimental cross, use the function est.map. The function plotMap, in addition to plotting a single map, can plot the comparison of two genetic maps (as long as they are composed of the same numbers of chromosomes and markers per chromosome). The function replace.map map be used to replace the genetic map in a cross with a new one.

newmap <- est.map(hyper, error.prob=0.01, verbose=TRUE)
plotMap(hyper, newmap)
hyper <- replace.map(hyper, newmap)

The function calc.errorlod may be used to assist in identifying possible genotyping errors; it calculates the error LOD scores described by Lincoln and Lander (1992). The calc.errorlod function return a modified version of the input cross, with error LOD scores included. The function top.errorlod prints the genotypes with values above a cutoff (by default, the cutoff is 4.0).

hyper <- calc.errorlod(hyper, error.prob=0.01)
top.errorlod(hyper)

The function plotGeno may be used to inspect the observed genotypes for a chromosome, with likely genotyping errors flagged.

plotGeno(hyper, chr=16, ind=c(24:34, 71:81))

Before doing QTL analyses, some intermediate calculations need to be performed. The function calc.genoprob calculates conditional genotype probabilities given the multipoint marker data. sim.geno simulates sequences of genotypes from their joint distribution, given the observed marker data.

As with calc.errorlod, these functions return a modified version of the input cross, with the intermediate calculations included. The step argument indicates the density of the grid on which the calculations will be performed, and determines the density at which LOD scores will be calculated.

hyper <- calc.genoprob(hyper, step=2.5, error.prob=0.01)
hyper <- sim.geno(hyper, step=2.5, n.draws=64, error.prob=0.01)

The function scanone performs a genome scan with a single QTL model. By default, it performs standard interval mapping (Lander and Botstein 1989): use of a normal model and the EM algorithm. If one specifies method="hk", Haley-Knott regression is performed (Haley and Knott 1992). These two methods require the results from calc.genoprob.

out.em <- scanone(hyper)
out.hk <- scanone(hyper, method="hk")

If one specifies method="imp", a genome scan is performed by the multiple imputation method of Sen and Churchill (2001). This method requires the results from sim.geno.

out.imp <- scanone(hyper, method="imp")

The output of scanone is a data.frame with class "scanone". The function plot.scanone may be used to plot the results, and may plot up to three sets of results against each other, as long as they conform appropriately.

plot(out.em)
plot(out.hk, col="blue", add=TRUE)
plot(out.imp, col="red", add=TRUE)
plot(out.hk, out.imp, out.em, chr=c(1,4), lty=1,
col=c("blue","red","black"))

The function summary.scanone may be used to list information on the peak LOD for each chromosome for which the LOD exceeds a specified threshold.

summary(out.em)
summary(out.em, threshold=3)
summary(out.hk, threshold=3)
summary(out.imp, threshold=3)

The function max.scanone returns the maximum LOD score, genome-wide.

max(out.em)
max(out.hk)
max(out.imp)

One may also use scanone to perform a permutation test to get a genome-wide LOD significance threshold.

operm.hk <- scanone(hyper, method="hk", n.perm=1000)

The result has class "scanoneperm". The summary.scanoneperm function may be used to calculate LOD thresholds.

summary(operm.hk, alpha=0.05)

The permutation results may also be used in the summary.scanone function to calculate LOD thresholds and genome-scan-adjusted p-values.

summary(out.hk, perms=operm.hk, alpha=0.05, pvalues=TRUE)

We should say at this point that the function save.image will save your workspace to disk. You'll wish you had used this if R crashes.

save.image()

The function scantwo performs a two-dimensional genome scan with a two-QTL model. Methods "em", "hk" and "imp" are all available. scantwo is considerably slower than scanone, and can require a great deal of memory. Thus, you may wish to re-run calc.genoprob and/or sim.geno with a more coarse grid.

hyper <- calc.genoprob(hyper, step=10, err=0.01)
hyper <- sim.geno(hyper, step=10, n.draws=64, err=0.01)

out2.hk <- scantwo(hyper, method="hk")
out2.em <- scantwo(hyper)
out2.imp <- scantwo(hyper, method="imp")

The output is an object with class scantwo. The function plot.scantwo may be used to plot the results. The upper triangle contains LOD scores for tests of epistasis, while the lower triangle contains LOD scores for the full model.

plot(out2.hk)
plot(out2.em)
plot(out2.imp)

The function summary.scantwo lists the interesting aspects of the output. For each pair of chromosomes $(k,l)$, it calculates the maximum LOD score for the full model, $M_f(k,l)$; a LOD score indicating evidence for a second QTL, allowing for epistasis), $M_{fv1}(k,l)$; a LOD score indicating evidence for epistasis, $M_i(k,l)$; the LOD score for the additive QTL model, $M_a(k,l)$; and a LOD score indicating evidence for a second QTL, assuming no epistasis, $M_{av1}(k,l)$.

You must provide five LOD thresholds, corresponding to the above five LOD scores, and in that order. A chromosome pair is printed if either (a) $M_f(k,l) \ge T_f$ and ($M_{fv1}(k,l) \ge T_{fv1}$ or $M_i(k,l) \ge T_i$), or (b) $M_a(k,l) \ge T_a$ and $M_{av1}(k,l) \ge T_{av1}$.

summary(out2.em, thresholds=c(6.2, 5.0, 4.6, 4.5, 2.3))
summary(out2.em, thresholds=c(6.2, 5.0, Inf, 4.5, 2.3))

In the latter case, the interaction LOD score will be ignored.

The function max.scantwo returns the maximum joint and additive LODs for a two-dimensional genome scan.

max(out2.em)

Permutation tests may also performed with scantwo; it may take a few days of CPU time. The output is a list containing the maxima of the above five LOD scores for each of the imputations.

operm2 <- scantwo(hyper, method="hk", n.perm=100)
summary(operm2, alpha=0.05)

Citing R/qtl

To cite R/qtl in publications, use the Broman et al. (2003) reference listed below.

Author(s)

Karl W Broman, [email protected]

References

Broman, K. W. and Sen, Ś. (2009) A guide to QTL mapping with R/qtl. Springer. https://rqtl.org/book/

Broman, K. W., Wu, H., Sen, Ś. and Churchill, G. A. (2003) R/qtl: QTL mapping in experimental crosses. Bioinformatics 19, 889–890.

Haley, C. S. and Knott, S. A. (1992) A simple regression method for mapping quantitative trait loci in line crosses using flanking markers. Heredity 69, 315–324.

Lander, E. S. and Botstein, D. (1989) Mapping Mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121, 185–199.

Lincoln, S. E. and Lander, E. S. (1992) Systematic detection of errors in genetic linkage data. Genomics 14, 604–610.

Sen, Ś. and Churchill, G. A. (2001) A statistical framework for quantitative trait mapping. Genetics 159, 371–387.

Description

Add dots at the locations of the selected marker covariates, for a plot of composite interval mapping results.

Usage

add.cim.covar(cimresult, chr, gap=25, ...)

Arguments

Details

One must first have used the function plot.scanone to plot the composite interval mapping results.

The arguments chr and gap must be identical to the values used in the call to plot.scanone.

Dots indicating the locations of the selected marker covariates are displayed on the x-axis. (By default, solid red circles are plotted; this may be modified by specifying the graphics parameters pch and col.)

Value

A data frame indicating the marker covariates that were plotted.

Author(s)

Karl W Broman, [email protected]

See Also

cim, plot.scanone

Examples

## Not run: data(hyper)
hyper <- calc.genoprob(hyper, step=2.5)

out <- scanone(hyper)
out.cim <- cim(hyper, n.marcovar=3)
plot(out, out.cim, chr=c(1,4,6,15), col=c("blue", "red"))

add.cim.covar(out.cim, chr=c(1,4,6,15))
## End(Not run)

Description

Add a significance threshold to a plot created by plot.scanone), using the permutation results.

Usage

add.threshold(out, chr, perms, alpha=0.05, lodcolumn=1, gap=25, ...)

Arguments

Details

This function allows you to add a horizontal line at the significance threshold to genome scan results plotted by plot.scanone.

The arguments out, chr, and gap must match what was used in the call to plot.scanone.

The argument perms must be specified. If X-chromosome-specific permutations were performed (via the argument perm.Xsp in the call to scanone), separate thresholds will be plotted for the autosomes and the X chromosome. These are calculated via the summary.scanoneperm function.

Value

None.

Author(s)

Karl W Broman, [email protected]

See Also

scanone, plot.scanone, summary.scanoneperm, xaxisloc.scanone

Examples

data(hyper)
hyper <- calc.genoprob(hyper)
out <- scanone(hyper, method="hk")
operm <- scanone(hyper, method="hk", n.perm=100, perm.Xsp=TRUE)

plot(out, chr=c(1,4,6,15,"X"))
add.threshold(out, chr=c(1,4,6,15,"X"), perms=operm, alpha=0.05)
add.threshold(out, chr=c(1,4,6,15,"X"), perms=operm, alpha=0.1,
              col="green", lty=2)

Description

Try adding all QTL x covariate interactions, one at a time, to a multiple QTL model, for a given set of covariates.

Usage

addcovarint(cross, pheno.col=1, qtl, covar=NULL, icovar, formula,
            method=c("imp","hk"), model=c("normal", "binary"),
            verbose=TRUE, pvalues=TRUE, simple=FALSE, tol=1e-4,
            maxit=1000, require.fullrank=FALSE)

Arguments

Details

The formula is used to specified the model to be fit. In the formula, use Q1, Q2, etc., or q1, q2, etc., to represent the QTLs, and the column names in the covariate data frame to represent the covariates.

We enforce a hierarchical structure on the model formula: if a QTL or covariate is in involved in an interaction, its main effect must also be included.

Value

An object of class addcovarint, with results as in the drop-one-term analysis from fitqtl. This is a data frame (given class "addcovarint", with the following columns: degrees of freedom (df), Type III sum of squares (Type III SS), LOD score(LOD), percentage of variance explained (%var), F statistics (F value), and P values for chi square (Pvalue(chi2)) and F distribution (Pvalue(F)).

Note that the degree of freedom, Type III sum of squares, the LOD score and the percentage of variance explained are the values comparing the full to the sub-model with the term dropped. Also note that for imputation method, the percentage of variance explained, the the F values and the P values are approximations calculated from the LOD score.

QTL x covariate interactions already included in the input formula are not tested.

Author(s)

Karl W Broman, [email protected]

References

Haley, C. S. and Knott, S. A. (1992) A simple regression method for mapping quantitative trait loci in line crosses using flanking markers. Heredity 69, 315–324.

Sen, Ś. and Churchill, G. A. (2001) A statistical framework for quantitative trait mapping. Genetics 159, 371–387.

See Also

addint, fitqtl, makeqtl, scanqtl, refineqtl, addqtl, addpair

Examples

data(fake.f2)

# take out several QTLs and make QTL object
qc <- c(1, 8, 13)
qp <- c(26, 56, 28)
fake.f2 <- subset(fake.f2, chr=qc)

fake.f2 <- calc.genoprob(fake.f2, step=2, err=0.001)
qtl <- makeqtl(fake.f2, qc, qp, what="prob")

# use the sex phenotype as the covariate
covar <- data.frame(sex=fake.f2$pheno$sex)

# try all possible QTL x sex interactions, one at a time
addcovarint(fake.f2, pheno.col=1, qtl, covar, "sex", y~Q1+Q2+Q3,
            method="hk")

Description

Try adding all possible pairwise interactions, one at a time, to a multiple QTL model.

Usage

addint(cross, pheno.col=1, qtl, covar=NULL, formula, method=c("imp","hk"),
       model=c("normal", "binary"), qtl.only=FALSE, verbose=TRUE,
       pvalues=TRUE, simple=FALSE, tol=1e-4, maxit=1000, require.fullrank=FALSE)

Arguments

Details

The formula is used to specified the model to be fit. In the formula, use Q1, Q2, etc., or q1, q2, etc., to represent the QTLs, and the column names in the covariate data frame to represent the covariates.

We enforce a hierarchical structure on the model formula: if a QTL or covariate is in involved in an interaction, its main effect must also be included.

Value

An object of class addint, with results as in the drop-one-term analysis from fitqtl. This is a data frame (given class "addint", with the following columns: degrees of freedom (df), Type III sum of squares (Type III SS), LOD score(LOD), percentage of variance explained (%var), F statistics (F value), and P values for chi square (Pvalue(chi2)) and F distribution (Pvalue(F)).

Note that the degree of freedom, Type III sum of squares, the LOD score and the percentage of variance explained are the values comparing the full to the sub-model with the term dropped. Also note that for imputation method, the percentage of variance explained, the the F values and the P values are approximations calculated from the LOD score.

Pairwise interactions already included in the input formula are not tested.

Author(s)

Karl W Broman, [email protected]

References

Haley, C. S. and Knott, S. A. (1992) A simple regression method for mapping quantitative trait loci in line crosses using flanking markers. Heredity 69, 315–324.

Sen, Ś. and Churchill, G. A. (2001) A statistical framework for quantitative trait mapping. Genetics 159, 371–387.

See Also

addcovarint, fitqtl, makeqtl, scanqtl, refineqtl, addqtl, addpair

Examples

data(fake.f2)

# take out several QTLs and make QTL object
qc <- c(1, 8, 13)
qp <- c(26, 56, 28)
fake.f2 <- subset(fake.f2, chr=qc)

fake.f2 <- calc.genoprob(fake.f2, step=2, err=0.001)
qtl <- makeqtl(fake.f2, qc, qp, what="prob")

# try all possible pairwise interactions, one at a time
addint(fake.f2, pheno.col=1, qtl, formula=y~Q1+Q2+Q3, method="hk")

Description

Add phenotype location(s) into a cross object (with eQTL/pQTL studies)

Usage

addloctocross(cross, locations=NULL, locfile="locations.txt", verbose=FALSE)

Arguments

Details

inputfile layout: Num Name Chr cM 1 X3.Hydroxypropyl 4 50.0 Num is the number of the phenotype in the cross object Name is the name of the phenotype (will be checked against the name already in the cross object at position num Chr Chromosome cM position from start of chromosome in cM

Value

The input cross object, with the locations added as an additional component locations

Author(s)

Ritsert C Jansen; Danny Arends; Pjotr Prins; Karl W Broman [email protected]

See Also

• mqmplot.cistrans - Cis/trans plot

• The MQM tutorial: https://rqtl.org/tutorials/MQM-tour.pdf

• MQM - MQM description and references

• mqmscan - Main MQM single trait analysis

• mqmscanall - Parallellized traits analysis

• mqmaugment - Augmentation routine for estimating missing data

• mqmautocofactors - Set cofactors using marker density

• mqmsetcofactors - Set cofactors at fixed locations

• mqmpermutation - Estimate significance levels

• scanone - Single QTL scanning

Examples

## Not run: 
    data(multitrait)
    data(locations)
    multiloc <- addloctocross(multitrait,locations)
    results <- scanall(multiloc)
    mqmplot.cistrans(results, multiloc, 5, FALSE, TRUE)
  
## End(Not run)

Description

Add a marker to a cross object.

Usage

addmarker(cross, genotypes, markername, chr, pos)

Arguments

Details

Use this function with caution. It would be best to incorporate new data into a single file to be imported with read.cross.

But if you have genotypes on one or two additional markers that you want to add, you might load them with read.csv and incorporate them with this function.

Value

The input cross object with the single marker added.

Author(s)

Karl W Broman, [email protected]

See Also

pull.markers, drop.markers

Examples

data(fake.f2)

# genotypes for new marker
gi <- pull.geno(fill.geno(fake.f2))[,"D5M197"]

# add marker to cross
fake.f2 <- addmarker(fake.f2, gi, "D5M197imp", "5", 11)

Description

Scan for an additional pair of QTL in the context of a multiple QTL model.

Usage

addpair(cross, chr, pheno.col=1, qtl, covar=NULL, formula,
        method=c("imp","hk"), model=c("normal", "binary"),
        incl.markers=FALSE, verbose=TRUE, tol=1e-4, maxit=1000,
        forceXcovar=FALSE)

Arguments

Details

The formula is used to specified the model to be fit. In the formula, use Q1, Q2, etc., or q1, q2, etc., to represent the QTLs, and the column names in the covariate data frame to represent the covariates.

We enforce a hierarchical structure on the model formula: if a QTL or covariate is in involved in an interaction, its main effect must also be included.

If neither of the two new QTL are indicated in the formula, we perform a two-dimensional scan as in scantwo. That is, for each pair of QTL positions, we fit two models: two additive QTL added to the formula, and two interacting QTL added to the formula.

If the both of the new QTL are indicated in the formula, that particular model is fit, with the positions of the new QTL allowed to vary across the genome. If just one of the QTL is indicated in the formula, a main effect for the other is added, and that particular model is fit, again with the positions of both QTL varying. Note that in this case the LOD scores are not analogous to those produced by scantwo. Thus, there slightly modified forms for the plots (produced by plot.scantwo) and summaries (produced by summary.scantwo and max.scantwo). In the plot, the x-axis is to be interpreted as the position of the first of the new QTL, and the y-axis is to be interpreted as the position of the second of the new QTL. In the summaries, we give the single best pair of positions on each pair of chromosomes, and give LOD scores comparing that pair of positions to the base model (without each of these QTL), and to the base model plus one additional QTL on one or the other of the chromosomes.

Value

An object of class scantwo, as produced by scantwo.

If neither of the new QTL were indicated in the formula, the result is just as in scantwo, though with LOD scores relative to the base model (omitting the new QTL).

Otherwise, the results are contained in what would ordinarily be in the full and additive LOD scores, with the additive LOD scores corresponding to the case that the first of the new QTL is to the left of the second of the new QTL, and the full LOD scores corresponding to the case that the first of the new QTL is to the right of the second of the new QTL. Because the structure of the LOD scores in this case is different from those output by scantwo, we include, in this case, an attribute "addpair"=TRUE. (We also require results of single-dimensional scans, omitting each of the two new QTL from the formula, one at a time; these are included as attributes "lod.minus1" and "lod.minus2".) The results are then treated somewhat differently by summary.scantwo, max.scantwo, and plot.scantwo. See the Details section.

Author(s)

Karl W Broman, [email protected]

References

Haley, C. S. and Knott, S. A. (1992) A simple regression method for mapping quantitative trait loci in line crosses using flanking markers. Heredity 69, 315–324.

Sen, Ś. and Churchill, G. A. (2001) A statistical framework for quantitative trait mapping. Genetics 159, 371–387.

See Also

addint, addqtl, fitqtl, makeqtl, scanqtl, refineqtl, makeqtl, scantwo, addtoqtl

Examples

# A totally contrived example to show some of what you can do

# simulate backcross data with 3 chromosomes (names "17", "18", "19")
#   one QTL on chr 17 at 40 cM
#   one QTL on chr 18 at 30 cM
#   two QTL on chr 19, at 10 and 40 cM
data(map10)
model <- rbind(c(1,40,0), c(2,30,0), c(3,10,0), c(3,40,0))
## Not run: fakebc <- sim.cross(map10[17:19], model=model, type="bc", n.ind=250)


# het at QTL on 17 and 1st QTL on 19 increases phenotype by 1 unit
# het at QTL on 18 and 2nd QTL on 19 decreases phenotype by 1 unit
qtlgeno <- fakebc$qtlgeno
phe <- rnorm(nind(fakebc))
w <- qtlgeno[,1]==2 & qtlgeno[,3]==2
phe[w] <- phe[w] + 1
w <- qtlgeno[,2]==2 & qtlgeno[,4]==2
phe[w] <- phe[w] - 1
fakebc$pheno[,1] <- phe

## Not run: fakebc <- calc.genoprob(fakebc, step=2, err=0.001)


# base model has QTLs on chr 17 and 18
qtl <- makeqtl(fakebc, chr=c("17", "18"), pos=c(40,30), what="prob")

# scan for an additional pair of QTL, one interacting with the locus
#     on 17 and one interacting with the locus on 18
out.ap <- addpair(fakebc, qtl=qtl, formula = y~Q1*Q3 + Q2*Q4, method="hk")

max(out.ap)
summary(out.ap)
plot(out.ap)

Description

Scan for an additional QTL in the context of a multiple QTL model.

Usage

addqtl(cross, chr, pheno.col=1, qtl, covar=NULL, formula,
       method=c("imp","hk"), model=c("normal", "binary"),
       incl.markers=TRUE, verbose=FALSE, tol=1e-4, maxit=1000,
       forceXcovar=FALSE, require.fullrank=FALSE)

Arguments

Details

The formula is used to specified the model to be fit. In the formula, use Q1, Q2, etc., or q1, q2, etc., to represent the QTLs, and the column names in the covariate data frame to represent the covariates.

We enforce a hierarchical structure on the model formula: if a QTL or covariate is in involved in an interaction, its main effect must also be included.

If one wishes to scan for QTL that interact with another QTL, include it in the formula (with an index of one more than the number of QTL in the input qtl object).

Value

An object of class scanone, as produced by the scanone function. LOD scores are relative to the base model (with any terms that include the new QTL omitted).

Author(s)

Karl W Broman, [email protected]

References

Haley, C. S. and Knott, S. A. (1992) A simple regression method for mapping quantitative trait loci in line crosses using flanking markers. Heredity 69, 315–324.

Sen, Ś. and Churchill, G. A. (2001) A statistical framework for quantitative trait mapping. Genetics 159, 371–387.

See Also

scanone, fitqtl, scanqtl, refineqtl, makeqtl, addtoqtl, addpair, addint

Examples

data(fake.f2)

# take out several QTLs and make QTL object
qc <- c(1, 8, 13)
qp <- c(26, 56, 28)

fake.f2 <- subset(fake.f2, chr=c(1,2,3,8,13))


fake.f2 <- calc.genoprob(fake.f2, step=2, err=0.001)
qtl <- makeqtl(fake.f2, qc, qp, what="prob")

# scan for an additional QTL
out1 <- addqtl(fake.f2, qtl=qtl, formula=y~Q1+Q2+Q3, method="hk")
max(out1)

# scan for an additional QTL that interacts with the locus on chr 1
out2 <- addqtl(fake.f2, qtl=qtl, formula=y~Q1*Q4+Q2+Q3, method="hk")
max(out2)

# plot interaction LOD scores
plot(out2-out1)

Description

Add a QTL or multiple QTL to a qtl object.

Usage

addtoqtl(cross, qtl, chr, pos, qtl.name, drop.lod.profile=TRUE)

Arguments

Value

An object of class qtl, just like the input qtl object, but with additional QTL added. See makeqtl for details.

Author(s)

Karl W Broman, [email protected]

See Also

makeqtl, fitqtl, dropfromqtl, replaceqtl, reorderqtl

Examples

data(fake.f2)

# take out several QTLs and make QTL object
qc <- c(1, 6, 13)
qp <- c(25.8, 33.6, 18.63)


fake.f2 <- calc.genoprob(fake.f2, step=2, err=0.001)
qtl <- makeqtl(fake.f2, qc, qp, what="prob")
qtl <- addtoqtl(fake.f2, qtl, 14, 35)

Description

In order to assess the support for a linkage group, this function splits the linkage groups into two pieces at each interval and in each case calculates a LOD score comparing the combined linkage group to the two pieces.

Usage

allchrsplits(cross, chr, error.prob=0.0001,
                map.function=c("haldane","kosambi","c-f","morgan"),
                m=0, p=0, maxit=4000, tol=1e-6, sex.sp=TRUE,
                verbose=TRUE)

Arguments

Value

A data frame (actually, an object of class "scanone", so that one may use plot.scanone, summary.scanone, etc.) with each row being an interval at which a split is made. The first two columns are the chromosome ID and midpoint of the interval. The third column is a LOD score comparing the combined linkage group to the split into two linkage groups. A fourth column (gap) indicates the length of each interval.

The row names indicate the flanking markers for each interval.

Author(s)

Karl W Broman, [email protected]

See Also

est.map, ripple, est.rf, switch.order, movemarker

Examples

data(fake.bc)
allchrsplits(fake.bc, 7, error.prob=0, verbose=FALSE)

Description

Uses the Viterbi algorithm to identify the most likely sequence of underlying genotypes, given the observed multipoint marker data, with possible allowance for genotyping errors.

Usage

argmax.geno(cross, step=0, off.end=0, error.prob=0.0001,
            map.function=c("haldane","kosambi","c-f","morgan"),
            stepwidth=c("fixed", "variable", "max"))

Arguments

Details

We use the Viterbi algorithm to calculate $\arg \max_v \Pr(g = v | O)$ where $g$ is the underlying sequence of genotypes and $O$ is the observed marker genotypes.

This is done by calculating $\gamma_k(v_k) = \max_{v_1, \ldots, v_{k-1}} \Pr(g_1 = v_1, \ldots, g_k = v_k, O_1, \ldots, O_k)$ for $k = 1, \ldots, n$ and then tracing back through the sequence.

Value

The input cross object is returned with a component, argmax, added to each component of cross$geno. The argmax component is a matrix of size [n.ind x n.pos], where n.pos is the number of positions at which the reconstructed genotypes were obtained, containing the most likely sequences of underlying genotypes. Attributes "error.prob", "step", and "off.end" are set to the values of the corresponding arguments, for later reference.

Warning

The Viterbi algorithm can behave badly when step is small but positive. One may observe quite different results for different values of step.

The problem is that, in the presence of data like A----H, the sequences AAAAAA and HHHHHH may be more likely than any one of the sequences AAAAAH, AAAAHH, AAAHHH, AAHHHH, AHHHHH, AAAAAH. The Viterbi algorithm produces a single "most likely" sequence of underlying genotypes.

Author(s)

Karl W Broman, [email protected]

References

Lange, K. (1999) Numerical analysis for statisticians. Springer-Verlag. Sec 23.3.

Rabiner, L. R. (1989) A tutorial on hidden Markov models and selected applications in speech recognition. Proceedings of the IEEE 77, 257–286.

See Also

sim.geno, calc.genoprob, fill.geno

Examples

data(fake.f2)
fake.f2 <- argmax.geno(fake.f2, step=2, off.end=5, err=0.01)

Description

Add or subtract LOD scores in results from scanone or scantwo.

Usage

scan1+scan2
scan1-scan2

Arguments

Details

This is used to calculate the sum or difference of LOD scores of two genome scan results. It is particularly useful for calculating the LOD scores for QTL-by-covariate interactions (see the example, below). Note that the degrees of freedom are also added or subtracted.

Value

The same type of data structure as the input objects, with LOD scores added or subtracted.

Author(s)

Karl W Broman, [email protected]

Examples

data(fake.bc)

fake.bc <- calc.genoprob(fake.bc, step=2.5)

# covariates
ac <- pull.pheno(fake.bc, c("sex","age"))
ic <- pull.pheno(fake.bc, "sex")

# scan with additive but not the interactive covariate
out.acovar <- scanone(fake.bc, addcovar=ac)

# scan with interactive covariate
out.icovar <- scanone(fake.bc, addcovar=ac, intcovar=ic)

# plot the difference of with and without the interactive covariate
#     This is a LOD score for a test of QTL x covariate interaction
plot(out.icovar-out.acovar)

Description

Add or subtract LOD scores in permutation results from scanone or scantwo.

Usage

perm1+perm2
perm1-perm2

Arguments

Details

This is used to calculate the sum or difference of LOD scores of two sets of permutation results from scanone or scantwo. One must be careful to ensure that the permutations are perfectly linked, which will require the use of set.seed.

Value

The same data structure as the input objects, with LOD scores added or subtracted.

Author(s)

Karl W Broman, [email protected]

Examples

data(fake.bc)

fake.bc <- calc.genoprob(fake.bc, step=2.5)

# covariates
ac <- pull.pheno(fake.bc, c("sex","age"))
ic <- pull.pheno(fake.bc, "sex")

# set seed
theseed <- round(runif(1, 1, 10^8))
set.seed(theseed)

# permutations with additive but not the interactive covariate
## Not run: operm.acovar <- scanone(fake.bc, addcovar=ac, n.perm=1000)


# re-set the seed
set.seed(theseed)

# permutations with interactive covariate
## Not run: operm.icovar <- scanone(fake.bc, addcovar=ac, intcovar=ic,
                      n.perm=1000)

## End(Not run)

# permutation results for the QTL x covariate interaction
operm.gxc <- operm.icovar - operm.acovar

# LOD thresholds
summary(operm.gxc)

Description

Simulated data for an intercross with some markers out of order.

Usage

data(badorder)

Format

An object of class cross. See read.cross for details.

Details

There are 250 F2 individuals typed at a total of 36 markers on four chromosomes. The data were simulated with QTLs at the center of chromosomes 1 and 3.

The order of several markers on chromosome 1 is incorrect. Markers on chromosomes 2 and 3 are switched.

Author(s)

Karl W Broman, [email protected]

See Also

est.rf, ripple, est.map, sim.cross

Examples

data(badorder)

# estimate recombination fractions
badorder <- est.rf(badorder)
plotRF(badorder)

# re-estimate map
newmap <- est.map(badorder)
plotMap(badorder, newmap)

# assess marker order on chr 1
rip3 <- ripple(badorder, chr=1, window=3)
summary(rip3)

Description

Calculate an approximate Bayesian credible interval for a particular chromosome, using output from scanone.

Usage

bayesint(results, chr, qtl.index, prob=0.95, lodcolumn=1, expandtomarkers=FALSE)

Arguments

Details

We take $10^{LOD}$, rescale it to have area 1, and then calculate the connected interval with density above some threshold and having coverage matching the target probability.

Value

An object of class scanone indicating the estimated QTL position and the approximate endpoints for the Bayesian credible interval.

Author(s)

Karl W Broman, [email protected]

See Also

scanone, lodint

Examples

data(hyper)

hyper <- calc.genoprob(hyper, step=0.5)
out <- scanone(hyper, method="hk")
bayesint(out, chr=1)
bayesint(out, chr=4)
bayesint(out, chr=4, prob=0.99)
bayesint(out, chr=4, expandtomarkers=TRUE)

Description

Data from bristle number in chromosome 3 recombinant isogenic lines of Drosophila melanogaster.

Usage

data(bristle3)

Format

An object of class cross. See read.cross for details.

Details

There are 66 chromosome 3 recombinant isogenic lines, derived from inbred lines that were selected for low (A) and high (B) abdominal bristle numbers. A recombinant chromosome 3 was placed in an isogenic low background.

There are eight phenotypes: the average and SD of the number of abdominal and sternopleural bristles in males and females for each line.

Each line is typed at 29 genetic markers on chromosome 3.

References

Long, A. D., Mullaney, S. L., Reid, L. A., Fry, J. D., Langley, C. H. and MacKay, T. F. C. (1995) High resolution mapping of genetic factors affecting abdominal bristle number in Drosophila melanogaster. Genetics 139, 1273–1291.

See Also

bristleX, listeria, fake.bc, fake.f2, fake.4way, hyper

Examples

data(bristle3)
# Summaries
summary(bristle3)
plot(bristle3)

# genome scan for each of the average phenotypes
bristle3 <- calc.genoprob(bristle3, step=2)
out <- scanone(bristle3, pheno.col=c(1,3,5,7))

# Plot the results
    # maximum LOD score among four phenotypes
ym <- max(apply(out[,-(1:2)], 2, max))
plot(out, lod=1:3, ylim=c(0,ym))
plot(out, lod=4, add=TRUE, col="green")

Description

Data from bristle number in chromosome X recombinant isogenic lines of Drosophila melanogaster.

Usage

data(bristleX)

Format

An object of class cross. See read.cross for details.

Details

There are 92 chromosome X recombinant isogenic lines, derived from inbred lines that were selected for low (A) and high (B) abdominal bristle numbers. A recombinant chromosome X was placed in an isogenic low background.

There are eight phenotypes: the average and SD of the number of abdominal and sternopleural bristles in males and females for each line.

Each line is typed at 17 genetic markers on chromosome 3.

References

Long, A. D., Mullaney, S. L., Reid, L. A., Fry, J. D., Langley, C. H. and MacKay, T. F. C. (1995) High resolution mapping of genetic factors affecting abdominal bristle number in Drosophila melanogaster. Genetics 139, 1273–1291.

See Also

bristleX, listeria, fake.bc, fake.f2, fake.4way, hyper

Examples

data(bristleX)
# Summaries
summary(bristleX)
plot(bristleX)

# genome scan for each of the average phenotypes
bristleX <- calc.genoprob(bristleX, step=2)
out <- scanone(bristleX, pheno.col=c(1,3,5,7))

# Plot the results
    # maximum LOD score among four phenotypes
ym <- max(apply(out[,-(1:2)], 2, max))
plot(out, lod=1:3, ylim=c(0,ym))
plot(out, lod=4, add=TRUE, col="green")

Description

Concatenate the data for multiple QTL experiments.

Usage

## S3 method for class 'cross'
c(...)

Arguments

Value

The concatenated input, as a cross object. Additional columns are added to the phenotype data indicating which cross an individual comes from; another column indicates cross type (0=BC, 1=intercross), if there are crosses of different types. The crosses are not required to have exactly the same set of phenotypes; phenotypes with the same names are assumed to be the same.

If the crosses have different sets of markers, we interpolate marker order, but the cM positions of markers that are in common between crosses must be precisely the same in the different crosses.

Author(s)

Karl W Broman, [email protected]

See Also

subset.cross

Examples

data(fake.f2)
junk <- fake.f2
junk <- c(fake.f2,junk)

Description

Concatenate the columns from different runs of scanone.

Usage

## S3 method for class 'scanone'
c(..., labels)
## S3 method for class 'scanone'
cbind(..., labels)

Arguments

Details

The aim of this function is to concatenate the results from multiple runs scanone, generally for different phenotypes and/or methods, to be used in parallel with summary.scanone.

Value

The concatenated input, as a scanone object.

Author(s)

Karl W Broman, [email protected]

See Also

summary.scanone, scanone, cbind.scanoneperm

Examples

data(fake.f2)
fake.f2 <- calc.genoprob(fake.f2)

out.hk <- scanone(fake.f2, method="hk")
out.np <- scanone(fake.f2, model="np")

out <- c(out.hk, out.np, labels=c("hk","np"))
plot(out, lod=1:2, col=c("blue", "red"))

Description

Concatenate the data for multiple runs of scanone with n.perm > 0.

Usage

## S3 method for class 'scanoneperm'
c(...)
## S3 method for class 'scanoneperm'
rbind(...)

Arguments

Details

The aim of this function is to concatenate the results from multiple runs of a permutation test scanone, to assist with the case that such permutations are done on multiple processors in parallel.

Value

The concatenated input, as a scanoneperm object.

Author(s)

Karl W Broman, [email protected]

See Also

summary.scanoneperm, scanone, cbind.scanoneperm, c.scantwoperm

Examples

data(fake.f2)

fake.f2 <- calc.genoprob(fake.f2)
operm1 <- scanone(fake.f2, method="hk", n.perm=100, perm.Xsp=TRUE)
operm2 <- scanone(fake.f2, method="hk", n.perm=50, perm.Xsp=TRUE)

operm <- c(operm1, operm2)

Description

Concatenate the columns from different runs of scantwo.

Usage

## S3 method for class 'scantwo'
c(...)
## S3 method for class 'scantwo'
cbind(...)

Arguments

Details

The aim of this function is to concatenate the results from multiple runs scantwo, generally for different phenotypes and/or methods.

Value

The concatenated input, as a scantwo object.

Author(s)

Karl W Broman, [email protected]

See Also

summary.scantwo, scantwo, c.scanone

Examples

data(fake.bc)
fake.bc <- calc.genoprob(fake.bc)

out2a <- scantwo(fake.bc, method="hk")
out2b <- scantwo(fake.bc, pheno.col=2, method="hk")

out2 <- c(out2a, out2b)

Description

Concatenate the data for multiple runs of scantwo with n.perm > 0.

Usage

## S3 method for class 'scantwoperm'
c(...)
## S3 method for class 'scantwoperm'
rbind(...)

Arguments

Details

The aim of this function is to concatenate the results from multiple runs of a permutation test scantwo, to assist with the case that such permutations are done on multiple processors in parallel.

Value

The concatenated input, as a scantwoperm object.

Author(s)

Karl W Broman, [email protected]

See Also

summary.scantwoperm, scantwo, cbind.scantwoperm

Examples

data(fake.f2)

fake.f2 <- calc.genoprob(fake.f2)
## Not run: operm1 <- scantwo(fake.f2, method="hk", n.perm=50)
operm2 <- scantwo(fake.f2, method="hk", n.perm=50)
## End(Not run)

operm <- c(operm1, operm2)

Description

Calculates a LOD score for each genotype, measuring the evidence for genotyping errors.

Usage

calc.errorlod(cross, error.prob=0.01,
              map.function=c("haldane","kosambi","c-f","morgan"),
              version=c("new","old"))

Arguments

Details

Calculates, for each individual at each marker, a LOD score measuring the strength of evidence for a genotyping error, as described by Lincoln and Lander (1992).

In the latest version, evidence for a genotype being in error is considered assuming that all other genotypes (for that individual, on that chromosome) are correct. The argument version allows one to specify whether this new version is used, or whether the original (old) version of the calculation is performed.

Note that values below 4 are generally not interesting. Also note that if markers are extremely tightly linked, recombination events can give large error LOD scores. The error LOD scores should not be trusted blindly, but should be viewed as a tool for identifying genotypes deserving further study.

Use top.errorlod to print all genotypes with error LOD scores above a specified threshold, plotErrorlod to plot the error LOD scores for specified chromosomes, and plotGeno to view the observed genotype data with likely errors flagged.

Value

The input cross object is returned with a component, errorlod, added to each component of cross$geno. The errorlod component is a matrix of size (n.ind x n.mar). An attribute "error.prob" is set to the value of the corresponding argument, for later reference.

Author(s)

Karl W Broman, [email protected]

References

Lincoln, S. E. and Lander, E. S. (1992) Systematic detection of errors in genetic linkage data. Genomics 14, 604–610.

See Also

plotErrorlod, top.errorlod, cleanGeno

Examples

data(hyper)

hyper <- calc.errorlod(hyper,error.prob=0.01)

# print those above a specified cutoff
top.errorlod(hyper, cutoff=4)

# plot genotype data, flagging genotypes with error LOD > cutoff
plotGeno(hyper, chr=1, ind=160:200, cutoff=7, min.sep=2)

Description

Uses the hidden Markov model technology to calculate the probabilities of the true underlying genotypes given the observed multipoint marker data, with possible allowance for genotyping errors.

Usage

calc.genoprob(cross, step=0, off.end=0, error.prob=0.0001,
              map.function=c("haldane","kosambi","c-f","morgan"),
              stepwidth=c("fixed", "variable", "max"))

Arguments

Details

Let $O_k$ denote the observed marker genotype at position $k$, and $g_k$ denote the corresponding true underlying genotype.

We use the forward-backward equations to calculate $\alpha_{kv} = \log Pr(O_1, \ldots, O_k, g_k = v)$ and $\beta_{kv} = \log Pr(O_{k+1}, \ldots, O_n | g_k = v)$

We then obtain $Pr(g_k | O_1, \ldots, O_n) = \exp(\alpha_{kv} + \beta_{kv}) / s$ where $s = \sum_v \exp(\alpha_{kv} + \beta_{kv})$

In the case of the 4-way cross, with a sex-specific map, we assume a constant ratio of female:male recombination rates within the inter-marker intervals.

Value

The input cross object is returned with a component, prob, added to each component of cross$geno. prob is an array of size [n.ind x n.pos x n.gen] where n.pos is the number of positions at which the probabilities were calculated and n.gen = 3 for an intercross, = 2 for a backcross, and = 4 for a 4-way cross. Attributes "error.prob", "step", "off.end", and "map.function" are set to the values of the corresponding arguments, for later reference (especially by the function calc.errorlod).

Author(s)

Karl W Broman, [email protected]

References

Lange, K. (1999) Numerical analysis for statisticians. Springer-Verlag. Sec 23.3.

Rabiner, L. R. (1989) A tutorial on hidden Markov models and selected applications in speech recognition. Proceedings of the IEEE 77, 257–286.

See Also

sim.geno, argmax.geno, calc.errorlod

Examples

data(fake.f2)
fake.f2 <- calc.genoprob(fake.f2, step=2, off.end=5)

data(fake.bc)
fake.bc <- calc.genoprob(fake.bc, step=0, off.end=0, err=0.01)

Description

Derive penalties for the penalized LOD scores (used by stepwiseqtl) on the basis of permutation results from a two-dimensional, two-QTL scan (obtained by scantwo).

Usage

calc.penalties(perms, alpha=0.05, lodcolumn)

Arguments

Details

Thresholds derived from scantwo permutations (that is, for a two-dimensional, two-QTL genome scan) are used to calculate penalties on main effects and interactions.

The main effect penalty is the 1-alpha quantile of the null distribution of the genome-wide maximum LOD score from a single-QTL genome scan (as with scanone).

The "heavy" interaction penalty is the 1-alpha quantile of the null distribution of the maximum interaction LOD score (that is, the $\log_{10}$ likelihood ratio comparing the best model with two interacting QTL to the best model with two additive QTL) from a two-dimensional, two-QTL genome scan (as with scantwo).

The "light" interaction penality is the difference between the "fv1" threshold from the scantwo permutations (that is, the 1-alpha quantile of the LOD score comparing the best model with two interacting QTL to the best single-QTL model) and the main effect penalty.

If the permutations results were obtained with perm.Xsp=TRUE, to give X-chr-specific results, six penalties are calculated: main effect for autosomes, main effect for X chr, heavy penalty on A:A interactions, light penalty on A:A interactions, penalty on A:X interactions, and penalty on X:X interactions.

Value

Vector of three values indicating the penalty on main effects and heavy and light penalties on interactions, or a matrix of such results, with each row corresponding to a different phenotype.

If the input permutations are X-chromosome-specific, the result has six values: main effect for autosomes, main effect for X chr, heavy penalty on A:A interactions, light penalty on A:A interactions, penalty on A:X interactions, and penalty on X:X interactions.

Author(s)

Karl W Broman, [email protected]

References

Manichaikul, A., Moon, J. Y., Sen, Ś, Yandell, B. S. and Broman, K. W. (2009) A model selection approach for the identification of quantitative trait loci in experimental crosses, allowing epistasis. Genetics, 181, 1077–1086.

See Also

scantwo, stepwiseqtl

Examples

data(fake.f2)

fake.f2 <- calc.genoprob(fake.f2, step=5)
out.2dim <- scantwo(fake.f2, method="hk")

# permutations

## Not run: permo.2dim <- scantwo(fake.f2, method="hk", n.perm=1000)
summary(permo.2dim, alpha=0.05)

# penalties
calc.penalties(permo.2dim)

Description

Concatenate the columns from different runs of scanone with n.perm > 0.

Usage

## S3 method for class 'scanoneperm'
cbind(..., labels)

Arguments

Details

The aim of this function is to concatenate the results from multiple runs of a permutation test scanone, generally for different phenotypes and/or methods, to be used in parallel with c.scanone.

Value

The concatenated input, as a scanoneperm object. If different numbers of permutation replicates were used, those columns with fewer replicates are padded with missing values (NA).

Author(s)

Karl W Broman, [email protected]

See Also

summary.scanoneperm, scanone, c.scanoneperm, c.scanone

Examples

data(fake.f2)
fake.f2 <- calc.genoprob(fake.f2)

operm1 <- scanone(fake.f2, method="hk", n.perm=10, perm.Xsp=TRUE)
operm2 <- scanone(fake.f2, method="em", n.perm=5, perm.Xsp=TRUE)

operm <- cbind(operm1, operm2, labels=c("hk","em"))
summary(operm)

Description

Column-bind permutations results from scantwo for multiple phenotypes or models.

Usage

## S3 method for class 'scantwoperm'
cbind(...)

Arguments

Value

The column-binded input, as a scantwoperm object.

Author(s)

Karl W Broman, [email protected]

See Also

scantwo, c.scantwoperm, summary.scantwoperm

Examples

data(fake.bc)

fake.bc <- calc.genoprob(fake.bc)
## Not run: operm1 <- scantwo(fake.bc, pheno.col=1, method="hk", n.perm=50)
operm2 <- scantwo(fake.bc, pheno.col=2, method="hk", n.perm=50)
## End(Not run)

operm <- cbind(operm1, operm2)

Description

Identify markers whose alleles might have been switched by comparing the LOD score for linkage to all other autosomal markers with the original data to that when the alleles have been switched.

Usage

checkAlleles(cross, threshold=3, verbose)

Arguments

Details

For each marker, we compare the maximum LOD score for the cases where the estimated recombination fraction > 0.5 to those where r.f. < 0.5. The function est.rf must first be run.

Note: Markers that are tightly linked to a marker whose alleles are switched are likely to also be flagged by this method. The real problem markers are likely those with the biggest difference in LOD scores.

Value

A data frame containing the flagged markers, having four columns: the marker name, chromosome ID, numeric index within chromosome, and the difference between the maximum two-point LOD score with the alleles switched to that from the original data.

Author(s)

Karl W Broman, [email protected]

See Also

est.rf, geno.crosstab, switchAlleles

Examples

data(fake.f2)


# switch homozygotes at marker D5M391
fake.f2 <- switchAlleles(fake.f2, "D5M391")

fake.f2 <- est.rf(fake.f2)
checkAlleles(fake.f2)

Description

Obtain the chromosome lengths in a cross or map object.

Usage

chrlen(object)

Arguments

Value

Returns a vector of chromosome lengths. If the cross has sex-specific maps, it returns a 2-row matrix with the two lengths for each chromosome.

Author(s)

Karl W Broman, [email protected]

See Also

summaryMap, pull.map, summary.cross

Examples

data(fake.f2)
chrlen(fake.f2)

map <- pull.map(fake.f2)
chrlen(map)

Description

Pull out the chromosome names from a cross object as one big vector.

Usage

chrnames(cross)

Arguments

Value

A vector of character strings (the chromosome names).

Author(s)

Karl W Broman, [email protected]

See Also

markernames, phenames

Examples

data(listeria)
chrnames(listeria)

Description

Composite interval mapping by a scheme from QTL Cartographer: forward selection at the markers (here, with filled-in genotype data) to a fixed number, followed by interval mapping with the selected markers as covariates, dropping marker covariates if they are within some fixed window size of the location under test.

Usage

cim(cross, pheno.col=1, n.marcovar=3, window=10,
    method=c("em", "imp", "hk", "ehk"),
    imp.method=c("imp", "argmax"), error.prob=0.0001,
    map.function=c("haldane", "kosambi", "c-v", "morgan"),
    addcovar=NULL, n.perm)

Arguments

Details

We first use fill.geno to impute any missing marker genotype data, either via a simple random imputation or using the Viterbi algorithm.

We then perform forward selection to a fixed number of markers. These will be used (again, with any missing data filled in) as covariates in the subsequent genome scan.

Value

The function returns an object of the same form as the function scanone:

If n.perm is missing, the function returns the scan results as a data.frame with three columns: chromosome, position, LOD score. Attributes indicate the names and positions of the chosen marker covariates.

If n.perm > 0, the function results the results of a permutation test: a vector giving the genome-wide maximum LOD score in each of the permutations.

Author(s)

Karl W Broman, [email protected]

References

Jansen, R. C. (1993) Interval mapping of multiple quantitative trait loci. Genetics, 135, 205–211.

Jansen, R. C. and Stam, P. (1994) High resolution of quantitative traits into multiple loci via interval mapping. Genetics, 136, 1447-1455.

Zeng, Z. B. (1993) Theoretical basis for separation of multiple linked gene effects in mapping quantitative trait loci. Proc. Natl. Acad. Sci. USA, 90, 10972–10976.

Zeng, Z. B. (1994) Precision mapping of quantitative trait loci. Genetics, 136, 1457–1468.

See Also

add.cim.covar, scanone, summary.scanone, plot.scanone, fill.geno

Examples

data(hyper)
hyper <- calc.genoprob(hyper, step=2.5)


out <- scanone(hyper)
out.cim <- cim(hyper, n.marcovar=3)
plot(out, out.cim, chr=c(1,4,6,15), col=c("blue", "red"))

add.cim.covar(out.cim, chr=c(1,4,6,15))

Description

Remove any intermediate calculations from a cross object.

Usage

## S3 method for class 'cross'
clean(object, ...)

Arguments

Value

The input object, with any intermediate calculations (such as is produced by calc.genoprob, argmax.geno and sim.geno) removed.

Author(s)

Karl W Broman, [email protected]

See Also

drop.nullmarkers, drop.markers, clean.scantwo

Examples

data(fake.f2)
names(fake.f2$geno)
fake.f2 <- calc.genoprob(fake.f2)
names(fake.f2$geno)
fake.f2 <- clean(fake.f2)
names(fake.f2$geno)

Description

In an object output from scantwo, replaces negative and missing LOD scores with 0, and replaces LOD scores for pairs of positions that are not separated by n.mar markers, or that are less than distance cM apart, with 0. Further, if the LOD for full model is less than the LOD for the additive model, the additive LOD is pasted over the full LOD.

Usage

## S3 method for class 'scantwo'
clean(object, n.mar=1, distance=0, ...)

Arguments

Value

The input scantwo object, with any negative or missing LOD scores replaced by 0, and LOD scores for pairs of positions separated by fewer than n.mar markers, or less than distance cM, are set to 0. Also, if the LOD for the full model is less than the LOD for the additive model, the additive LOD is used in place of the full LOD.

Author(s)

Karl W Broman, [email protected]

See Also

scantwo, summary.scantwo

Examples

data(fake.f2)

fake.f2 <- calc.genoprob(fake.f2, step=5)
out2 <- scantwo(fake.f2, method="hk")
out2 <- clean(out2)
out2cl2 <- clean(out2, n.mar=2, distance=5)

Description

Delete genotypes from a cross that are indicated to be possibly in error, as they result in apparent tight double-crossovers.

Usage

cleanGeno(cross, chr, maxdist=2.5, maxmark=2, verbose=TRUE)

Arguments

Details

We first use locateXO to identify crossover locations. If a pair of adjacted crossovers are separated by no more than maxdist and contain no more than maxmark genotyped markers, the intervening genotypes are omitted (that is, changed to NA).

The arguments maxdist and maxmark may be vectors. (If both have length greater than 1, they must have the same length.) If they are vectors, genotypes are omitted if they satisify any one of the (maxdist, maxmark) pairs.

Value

The input cross object with suspect genotypes omitted.

Author(s)

Karl W Broman, [email protected]

See Also

locateXO, countXO, calc.errorlod

Examples

data(hyper)
sum(ntyped(hyper))
hyperc <- cleanGeno(hyper, chr=4, maxdist=c(2.5, 10), maxmark=c(2, 1))
sum(ntyped(hyperc))

Description

Verify that two objects of class cross have identical classes, chromosomes, markers, genotypes, genetic maps, and phenotypes.

Usage

comparecrosses(cross1, cross2, tol=1e-5)

Arguments

Value

None.

Author(s)

Karl W Broman, [email protected]

See Also

summary.cross

Examples

data(listeria)
comparecrosses(listeria, listeria)

Description

Count proportion of matching genotypes between all pairs of individuals, to look for unusually closely related individuals.

Usage

comparegeno(cross, what=c("proportion","number","both"))

Arguments

Value

A matrix whose (i,j)th element is the proportion or number of matching genotypes for individuals i and j.

If called with what="both", the lower triangle contains the proportion and the upper triangle contains the number.

If called with what="proportion", the diagonal contains missing values. Otherwise, the diagonal contains the number of typed markers for each individual.

The output is given class "comparegeno" so that appropriate summary and plot functions may be used.

Author(s)

Karl W Broman, [email protected]

See Also

nmissing, summary.comparegeno, plot.comparegeno

Examples

data(listeria)

cg <- comparegeno(listeria)

summary(cg, 0.7)
plot(cg)

Description

Compare the likelihood of an alternative order for markers on a chromosome to the current order.

Usage

compareorder(cross, chr, order, error.prob=0.0001,
             map.function=c("haldane","kosambi","c-f","morgan"),
             maxit=4000, tol=1e-6, sex.sp=TRUE)

Arguments

Value

A data frame with two rows: the current order in the input cross object, and the revised order. The first column is the log10 likelihood of the new order relative to the original one (positive values indicate that the new order is better supported). The second column is the estimated genetic length of the chromosome for each order. In the case of sex-specific maps, there are separate columns for the female and male genetic lengths.

Author(s)

Karl W Broman, [email protected]

See Also

ripple, switch.order, movemarker

Examples

data(badorder)
compareorder(badorder, chr=1, order=c(1:8,11,10,9,12))

Description

Produces a very condensed version of the output of scantwo.

Usage

## S3 method for class 'scantwo'
condense(object)

Arguments

Details

This produces a very reduced version of the output of scantwo, for which a summary may still be created via summary.scantwo, though plots can no longer be made.

Value

An object of class scantwocondensed, containing just the maximum full, additive and interactive LOD scores, and the positions where they occured, on each pair of chromosomes.

Author(s)

Karl W Broman, [email protected]

See Also

scantwo, summary.scantwo, max.scantwo

Examples

data(fake.f2)

fake.f2 <- calc.genoprob(fake.f2)

out2 <- scantwo(fake.f2, method="hk")

out2c <- condense(out2)
summary(out2c, allpairs=FALSE)
max(out2c)

Description

Convert a genetic map from using one map function to another.

Usage

## S3 method for class 'map'
convert(object, old.map.function=c("haldane", "kosambi", "c-f", "morgan"),
         new.map.function=c("haldane", "kosambi", "c-f", "morgan"), ...)

Arguments

Details

The location of the first marker on each chromosome is left unchanged. Inter-marker distances are converted to recombination fractions with the inverse of the old.map.function, and then back to distances with the new.map.function.

Value

The same as the input, but with inter-marker distances changed to reflect a different map function.

Author(s)

Karl W Broman, [email protected]

See Also

est.map, replace.map

Examples

data(listeria)
map <- pull.map(listeria)
map <- convert(map, "haldane", "kosambi")
listeria <- replace.map(listeria, map)

Description

Convert the output from scanone from the format used in R/qtl version 0.97 and earlier to that used in version 0.98 and later.

Usage

## S3 method for class 'scanone'
convert(object, ...)

Arguments

Details

Previously, inter-marker locations were named as, for example, loc7.5.c3; these were changed to c3.loc7.5.

Value

The same scanone output, but revised for use with R/qtl version 0.98 and later.

Author(s)

Karl W Broman, [email protected]

See Also

scanone, convert.scantwo

Examples

## Not run: out.new <- convert(out.old)

Description

Convert the output from scantwo from the format used in R/qtl version 1.03 and earlier to that used in version 1.04 and later.

Usage

## S3 method for class 'scantwo'
convert(object, ...)

Arguments

Details

Previously, the output from scantwo contained the full and interaction LOD scores. In R/qtl version 1.04 and later, the output contains the LOD scores from the full and additive QTL models.

Value

The same scanone output, but revised for use with R/qtl version 1.03 and later.

Author(s)

Karl W Broman, [email protected]

See Also

scantwo, convert.scanone

Examples

## Not run: out2.new <- convert(out2.old)

Description

Convert a cross to type "riself" (RIL by selfing).

Usage

convert2riself(cross)

Arguments

Details

If there are more genotypes with code 3 (BB) than code 2 (AB), we omit the genotypes with code==2 and call those with code==3 the BB genotypes.

If, instead, there are more genotypes with code 2 than code 3, we omit the genotypes with code==3 and call those with code==2 the BB genotypes.

Any chromosomes with class "X" (X chromosome) are changed to class "A" (autosomal).

Value

The input cross object, with genotype codes possibly changed and cross type changed to "riself".

Author(s)

Karl W Broman, [email protected]

See Also

convert2risib

Examples

data(hyper)
hyper.as.riself <- convert2riself(hyper)

Description

Convert a cross to type "risib" (RIL by sib mating).

Usage

convert2risib(cross)

Arguments

Details

If there are more genotypes with code 3 (BB) than code 2 (AB), we omit the genotypes with code==2 and call those with code==3 the BB genotypes.

If, instead, there are more genotypes with code 2 than code 3, we omit the genotypes with code==3 and call those with code==2 the BB genotypes.

Value

The input cross object, with genotype codes possibly changed and cross type changed to "risib".

Author(s)

Karl W Broman, [email protected]

See Also

convert2riself

Examples

data(hyper)
hyper.as.risib <- convert2risib(hyper)

Description

Convert a sex-specific map to a sex-averaged one, assuming that the female and male maps are actually the same (that is, that the map was estimated assuming a common recombination rate in females and males).

Usage

convert2sa(map, tol=1e-4)

Arguments

Details

We pull out just the female marker locations, and give a warning if there are large differences between the female and male maps.

Value

A map object, with sex-averaged distances.

Author(s)

Karl W Broman, [email protected]

See Also

est.map, plotMap

Examples

data(fake.4way)
## Not run: fake.4way <- subset(fake.4way, chr="-X")

nm <- est.map(fake.4way, sex.sp=FALSE)
plot(convert2sa(nm))

Description

Count the number of obligate crossovers for each individual in a cross, either by chromosome or overall.

Usage

countXO(cross, chr, bychr=FALSE)

Arguments

Details

For each individual we count the minimal number of crossovers that explain the observed genotype data.

Value

If bychr=TRUE, a matrix of counts is returned, with rows corresponding to individuals and columns corresponding to chromosomes.

If bychr=FALSE, a vector of counts (the total number of crossovers across all selected chromosomes) is returned.

Author(s)

Karl W Broman, [email protected]

See Also

ripple, locateXO, cleanGeno

Examples

data(hyper)
plot(countXO(hyper))

Description

Drop markers with duplicate names; retaining the first of each set, with consensus genotyps

Usage

drop.dupmarkers(cross, verbose=TRUE)

Arguments

Value

The input cross object, with any duplicate markers omitted (except for one). The marker retained will have consensus genotypes; if multiple versions of a marker have different genotypes for an individual, they will be replaced by NA.

Any derived data (such as produced by calc.genoprob) will be stripped off.

Author(s)

Karl W Broman, [email protected]

See Also

drop.nullmarkers, pull.markers, drop.markers, summary.cross, clean.cross

Examples

data(listeria)

listeria <- drop.dupmarkers(listeria)

Description

Drop a vector of markers from the data matrices and genetic maps.

Usage

drop.markers(cross, markers)

Arguments

Value

The input object, with any markers in the vector markers removed from the genotype data matrices, genetic maps, and, if applicable, any derived data (such as produced by calc.genoprob). (It might be a good idea to re-derive such things after using this function.)

Author(s)

Karl W Broman, [email protected]

See Also

drop.nullmarkers, pull.markers, geno.table, clean.cross

Examples

data(listeria)
listeria2 <- drop.markers(listeria, c("D10M44","D1M3","D1M75"))

Description

Drop markers, from the data matrices and genetic maps, that have no genotype data.

Usage

drop.nullmarkers(cross)

Arguments

Value

The input object, with any markers lacking genotype data removed from the genotype data matrices, genetic maps, and, if applicable, any derived data (such as produced by calc.genoprob). (It might be a good idea to re-derive such things after using this function.)

Author(s)

Karl W Broman, [email protected]

See Also

nullmarkers, drop.markers, clean.cross, geno.table

Examples

# removes one marker from hyper
data(hyper)
hyper <- drop.nullmarkers(hyper)

# shouldn't do anything to listeria
data(listeria)
listeria <- drop.nullmarkers(listeria)

Description

Drop a QTL or multiple QTL from a QTL object

Usage

dropfromqtl(qtl, index, chr, pos, qtl.name, drop.lod.profile=TRUE)

Arguments

Details

Provide either chr and pos, or one of qtl.name or index.

Value

The input qtl object with the specified QTL omitted. See makeqtl for details on the format.

Author(s)

Karl W Broman, [email protected]

See Also

makeqtl, fitqtl, addtoqtl, replaceqtl , reorderqtl

Examples

data(fake.f2)

# take out several QTLs and make QTL object
qc <- c(1, 6, 13)
qp <- c(25.8, 33.6, 18.63)
fake.f2 <- subset(fake.f2, chr=qc)

fake.f2 <- calc.genoprob(fake.f2, step=2, err=0.001)
qtl <- makeqtl(fake.f2, qc, qp, what="prob")

newqtl <- dropfromqtl(qtl, chr=1, pos=25.8)
altqtl <- dropfromqtl(qtl, index=1)

Description

Drop one marker at a time from a genetic map and calculate the change in log likelihood and in the chromosome length, in order to identify problematic markers.

Usage

droponemarker(cross, chr, error.prob=0.0001,
                map.function=c("haldane","kosambi","c-f","morgan"),
                m=0, p=0, maxit=4000, tol=1e-6, sex.sp=TRUE,
                verbose=TRUE)

Arguments

Value

A data frame (actually, an object of class "scanone", so that one may use plot.scanone, summary.scanone, etc.) with each row being a marker. The first two columns are the chromosome ID and position. The third column is a LOD score comparing the hypothesis that the marker is not linked to the hypothesis that it belongs at that position.

In the case of a 4-way cross, with sex.sp=TRUE, there are two additional columns with the change in the estimated female and male genetic lengths of the respective chromosome, upon deleting that marker. With sex.sp=FALSE, or for other types of crosses, there is one additional column, with the change in estimated genetic length of the respective chromosome, when the marker is omitted.

A well behaved marker will have a negative LOD score and a small change in estimated genetic length. A poorly behaved marker will have a large positive LOD score and a large change in estimated genetic length. But note that dropping the first or last marker on a chromosome could result in a large change in estimated length, even if they are not badly behaved; for these markers one should focus on the LOD scores, with a large positive LOD score being bad.

Author(s)

Karl W Broman, [email protected]

See Also

tryallpositions, est.map, ripple, est.rf, switch.order, movemarker, drop.markers

Examples

data(fake.bc)
droponemarker(fake.bc, 7, error.prob=0, verbose=FALSE)

Description

Plot the phenotype means for each group defined by the genotypes at one or two markers (or the values at a discrete covariate).

Usage

effectplot(cross, pheno.col=1, mname1, mark1, geno1, mname2, mark2,
           geno2, main, ylim, xlab, ylab, col, add.legend=TRUE,
           legend.lab, draw=TRUE, var.flag=c("pooled","group"))

Arguments

Details

In the plot, the y-axis is the phenotype. In the case of one marker, the x-axis is the genotype for that marker. In the case of two markers, the x-axis is for different genotypes of the second marker, and the genotypes of first marker are represented by lines in different colors. Error bars are plotted at $\pm$ 1 SE.

The results of sim.geno are used; if they are not available, sim.geno is run with n.draws=16. The average phenotype for each genotype group takes account of missing genotype data by averaging across the imputations. The SEs take account of both the residual phenotype variation and the imputation error.

Value

A data.frame containing the phenotype means and standard errors for each group.

Author(s)

Hao Wu; Karl W Broman, [email protected]

See Also

plotPXG, find.marker, effectscan, find.pseudomarker

Examples

data(fake.f2)


# impute genotype data
## Not run: fake.f2 <- sim.geno(fake.f2, step=5, n.draws=64)


########################################
# one marker plots
########################################
### plot of genotype-specific phenotype means for 1 marker
mname <- find.marker(fake.f2, 1, 37) # marker D1M437
effectplot(fake.f2, pheno.col=1, mname1=mname)

### output of the function contains the means and SEs
output <- effectplot(fake.f2, mname1=mname)
output

### plot a phenotype
# Plot of sex-specific phenotype means,
# note that "sex" must be a phenotype name here
effectplot(fake.f2, mname1="sex", geno1=c("F","M"))
# alternatively:
sex <- pull.pheno(fake.f2, "sex")
effectplot(fake.f2, mname1="Sex", mark1=sex, geno1=c("F","M"))

########################################
# two markers plots
########################################

### plot two markers
# plot of genotype-specific phenotype means for 2 markers
mname1 <- find.marker(fake.f2, 1, 37) # marker D1M437
mname2 <- find.marker(fake.f2, 13, 24) # marker D13M254
effectplot(fake.f2, mname1=mname1, mname2=mname2)

### plot two pseudomarkers
#####  refer to pseudomarkers by their positions
effectplot(fake.f2, mname1="1@35", mname2="13@25")

#####  alternatively, find their names via find.pseudomarker
pmnames <- find.pseudomarker(fake.f2, chr=c(1, 13), c(35, 25))
effectplot(fake.f2, mname1=pmnames[1], mname2=pmnames[2])

### Plot of sex- and genotype-specific phenotype means
mname <- find.marker(fake.f2, 13, 24) # marker D13M254
# sex and a marker
effectplot(fake.f2, mname1=mname, mname2="Sex",
           mark2=sex, geno2=c("F","M"))

# Same as above, switch role of sex and the marker
# sex and marker
effectplot(fake.f2, mname1="Sex", mark1=sex,
           geno1=c("F","M"), mname2=mname)

# X chromosome marker
mname <- find.marker(fake.f2, "X", 14) # marker DXM66
effectplot(fake.f2, mname1=mname)

# Two markers, including one on the X
mnames <- find.marker(fake.f2, c(13, "X"), c(24, 14))
effectplot(fake.f2, mname1=mnames[1], mname2=mnames[2])

Description

This function is used to plot the estimated QTL effects along selected chromosomes. For a backcross, there will be only one line, representing the additive effect. For an intercross, there will be two lines, representing the additive and dominance effects.

Usage

effectscan(cross, pheno.col=1, chr, get.se=FALSE, draw=TRUE,
           gap=25, ylim, mtick=c("line","triangle"),
           add.legend=TRUE, alternate.chrid=FALSE, ...)

Arguments

Details

The results of sim.geno are required for taking account of missing genotype information.

For a backcross, the additive effect is estimated as the difference between the phenotypic averages for heterozygotes and homozygotes.

For recombinant inbred lines, the additive effect is estimated as half the difference between the phenotypic averages for the two homozygotes.

For an intercross, the additive and dominance effects are estimated from linear regression on $a$ and $d$ with $a$ = -1, 0, 1, for the AA, AB and BB genotypes, respectively, and $d$ = 0, 1, 0, for the AA, AB and BB genotypes, respectively.

As usual, the X chromosome is a bit more complicated. We estimate separate additive effects for the two sexes, and for the two directions within females.

There is an internal function plot.effectscan that creates the actual plot by calling plot.scanone. In the case get.se=TRUE, colored regions indicate $\pm$ 1 SE.

Value

The results are returned silently, as an object of class "effectscan", which is the same as the form returned by the function scanone, though with estimated effects where LOD scores might be. That is, it is a data frame with the first two columns being chromosome ID and position (in cM), and subsequent columns being estimated effects, and (if get.se=TRUE) standard errors.

Author(s)

Karl W. Broman, [email protected]

References

Sen, Ś. and Churchill, G. A. (2001) A statistical framework for quantitative trait mapping. Genetics 159, 371–387.

See Also

effectplot, plotPXG, sim.geno

Examples

data(fake.f2)

fake.f2 <- sim.geno(fake.f2, step=2.5, n.draws=16)

# allelic effect on whole genome
effectscan(fake.f2)

# on chromosome 13, include standard errors
effectscan(fake.f2, chr="13", mtick="triangle", get.se=TRUE)

Description

Uses the Lander-Green algorithm (i.e., the hidden Markov model technology) to re-estimate the genetic map for an experimental cross.

Usage

est.map(cross, chr, error.prob=0.0001,
        map.function=c("haldane","kosambi","c-f","morgan"),
        m=0, p=0, maxit=10000, tol=1e-6, sex.sp=TRUE,
        verbose=FALSE, omit.noninformative=TRUE, offset, n.cluster=1)

Arguments

Details

By default, the map is estimated assuming no crossover interference, but a map function is used to derive the genetic distances (though, by default, the Haldane map function is used).

For a backcross or intercross, inter-marker distances may be estimated using the Stahl model for crossover interference, of which the chi-square model is a special case.

In the chi-square model, points are tossed down onto the four-strand bundle according to a Poisson process, and every $(m+1)$st point is a chiasma. With the assumption of no chromatid interference, crossover locations on a random meiotic product are obtained by thinning the chiasma process. The parameter $m$ (a non-negative integer) governs the strength of crossover interference, with $m=0$ corresponding to no interference.

In the Stahl model, chiasmata on the four-strand bundle are a superposition of chiasmata from two mechanisms, one following a chi-square model and one exhibiting no interference. An additional parameter, $p$, gives the proportion of chiasmata from the no interference mechanism.

Value

A map object; a list whose components (corresponding to chromosomes) are either vectors of marker positions (in cM) or matrices with two rows of sex-specific marker positions. The maximized log likelihood for each chromosome is saved as an attribute named loglik. In the case that estimation was under an interference model (with m > 0), allowed only for a backcross, m and p are also included as attributes.

Author(s)

Karl W Broman, [email protected]

References

Armstrong, N. J., McPeek, M. J. and Speed, T. P. (2006) Incorporating interference into linkage analysis for experimental crosses. Biostatistics 7, 374–386.

Lander, E. S. and Green, P. (1987) Construction of multilocus genetic linkage maps in humans. Proc. Natl. Acad. Sci. USA 84, 2363–2367.

Lange, K. (1999) Numerical analysis for statisticians. Springer-Verlag. Sec 23.3.

Rabiner, L. R. (1989) A tutorial on hidden Markov models and selected applications in speech recognition. Proceedings of the IEEE 77, 257–286.

Zhao, H., Speed, T. P. and McPeek, M. S. (1995) Statistical analysis of crossover interference using the chi-square model. Genetics 139, 1045–1056.

See Also

map2table, plotMap, replace.map, est.rf, fitstahl

Examples

data(fake.f2)

newmap <- est.map(fake.f2)
logliks <- sapply(newmap, attr, "loglik")
plotMap(fake.f2, newmap)
fake.f2 <- replace.map(fake.f2, newmap)

Description

Estimate the sex-averaged recombination fraction between all pairs of genetic markers.

Usage

est.rf(cross, maxit=10000, tol=1e-6)

Arguments

Details

For a backcross, one can simply count recombination events. For an intercross or 4-way cross, a version of the EM algorithm must be used to estimate recombination fractions. (Since, for example, in an intercross individual that is heterozygous at two loci, it is not known whether there were 0 or 2 recombination events.) Note that, for the 4-way cross, we estimate sex-averaged recombination fractions.

Value

The input cross object is returned with a component, rf, added. This is a matrix of size (tot.mar x tot.mar). The diagonal contains the number of typed meioses per marker, the lower triangle contains the estimated recombination fractions, and the upper triangle contains the LOD scores (testing rf = 0.5).

Author(s)

Karl W Broman, [email protected]

See Also

plotRF, pull.rf, plot.rfmatrix, est.map, badorder, checkAlleles

Examples

data(badorder)
badorder <- est.rf(badorder)
plotRF(badorder)

Description

Simulated data for a phase-known 4-way cross, obtained using sim.cross.

Usage

data(fake.4way)

Format

An object of class cross. See read.cross for details.

Details

There are 250 individuals typed at 157 markers, including 8 on the X chromosome.

There are two phenotypes (including sex, for which 0=female and 1=male). The quantitative phenotype is affected by three QTLs: two on chromosome 2 at positions 10 and 25 cM on the female genetic map, and one on chromosome 7 at position 40 cM on the female map.

Author(s)

Karl W Broman, [email protected]

See Also

sim.cross, fake.bc, fake.f2, listeria, hyper, bristle3, bristleX

Examples

data(fake.4way)

plot(fake.4way)
summary(fake.4way)

# estimate recombination fractions
fake.4way <- est.rf(fake.4way)
plotRF(fake.4way)

# estimate genetic maps
ssmap <- est.map(fake.4way, verbose=TRUE)
samap <- est.map(fake.4way, sex.sp=FALSE, verbose=TRUE)
plot(ssmap, samap)

# error lod scores
fake.4way <- calc.genoprob(fake.4way, err=0.01)
fake.4way <- calc.errorlod(fake.4way, err=0.01)
top.errorlod(fake.4way, cutoff=2.5)

# genome scan
fake.4way <- calc.genoprob(fake.4way, step=2.5)
out.hk <- scanone(fake.4way, method="hk")
out.em <- scanone(fake.4way, method="em")
plot(out.em,out.hk,chr=c(2,7))

Description

Simulated data for a backcross, obtained using sim.cross.

Usage

data(fake.bc)

Format

An object of class cross. See read.cross for details.

Details

There are 400 backcross individuals typed at 91 markers and with two phenotypes and two covariates (sex and age).

The two phenotypes are due to four QTLs, with no epistasis. There is one on chromosome 2 (at 30 cM), two on chromosome 5 (at 10 and 50 cM), and one on chromosome 10 (at 30 cM). The QTL on chromosome 2 has an effect only in the males (sex=1); the two QTLs on chromosome 5 have effect in coupling for the first phenotype and in repulsion for the second phenotype. Age has an effect of increasing the phenotypes.

Author(s)