plot.info: Plot the proportion of missing genotype information

Description

Plot a measure of the proportion of missing information in the genotype data.

Usage

plot.info(x, chr, method=c("both","entropy","variance"), step=1,
          off.end=0, error.prob=0.001,
          map.function=c("haldane","kosambi","c-f","morgan"),
          alternate.chrid=FALSE, ...)

Arguments

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

chr

Optional vector indicating the chromosomes to plot. 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

method

Indicates whether to plot the entropy version of the information, the variance version, or both.

step

Maximum distance (in cM) between positions at which the missing information is calculated, though for step=0, it is are calculated only at the marker locations.

off.end

Distance (in cM) past the terminal markers on each chromosome to which the genotype probability calculations will be carried.

error.prob

Assumed genotyping error rate used in the calculation of the penetrance Pr(observed genotype | true genotype).

map.function

Indicates whether to use the Haldane, Kosambi or Carter-Falconer map function when converting genetic distances into recombination fractions.

alternate.chrid

If TRUE and more than one chromosome is plotted, alternate the placement of chromosome axis labels, so that they may be more easily distinguished.

...

Passed to plot.scanone.

Value

An object with class scanone: a data.frame with columns the chromosome IDs and cM positions followed by the entropy and/or variance version of the missing information.

Details

The entropy version of the missing information: for a single individual at a single genomic position, we measure the missing information as $H = \sum_g p_g \log p_g / \log n$, where $p_g$ is the probability of the genotype $g$, and $n$ is the number of possible genotypes, defining $0 \log 0 = 0$. This takes values between 0 and 1, assuming the value 1 when the genotypes (given the marker data) are equally likely and 0 when the genotypes are completely determined. We calculate the missing information at a particular position as the average of $H$ across individuals. For an intercross, we don't scale by $\log n$ but by the entropy in the case of genotype probabilities (1/4, 1/2, 1/4). The variance version of the missing information: we calculate the average, across individuals, of the variance of the genotype distribution (conditional on the observed marker data) at a particular locus, and scale by the maximum such variance.

Calculations are done in C (for the sake of speed in the presence of little thought about programming efficiency) and the plot is created by a call to plot.scanone.

Note that summary.scanone may be used to display the maximum missing information on each chromosome.

Examples

Run this code

data(hyper)
hyper <- subset(hyper,chr=1:4)
plot.info(hyper,chr=c(1,4))

# save the results and view maximum missing info on each chr
info <- plot.info(hyper)
summary(info)