1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202
|
\name{summary.scanone}
\alias{summary.scanone}
\title{Summarize the results of a genome scans}
\description{
Print the rows of the output from \code{\link{scanone}} that
correspond to the maximum LOD for each chromosome, provided that they
exceed some specified thresholds.
}
\usage{
\method{summary}{scanone}(object, threshold,
format=c("onepheno", "allpheno", "allpeaks", "tabByCol", "tabByChr"),
perms, alpha, lodcolumn=1, pvalues=FALSE,
ci.function=c("lodint", "bayesint"), \dots)
}
\arguments{
\item{object}{An object output by the function
\code{\link{scanone}}.}
\item{threshold}{LOD score thresholds. Only peaks with LOD score above
this value will be returned. This could be a single number or (for
formats other than \code{"onepheno"}) a threshold for each LOD score
column. If \code{alpha} is specified, \code{threshold} should not be.}
\item{format}{Format for the output. See Details, below.}
\item{perms}{Optional permutation results used to derive thresholds or
to calculate genome-scan-adjusted p-values. This must be consistent
with the \code{object} input, in that it must have the same number of
LOD score columns, though it can have just one column of permutation
results, in which case they are reused for all LOD score columns in
the \code{\link{scanone}} output, \code{object}.
(These can also be permutation results from
\code{\link{scantwo}}, which permutations for a one-dimensional scan.)}
\item{alpha}{If perms are included, this is the significance level used
to calculate thresholds for determining which peaks to pull out.
If \code{threshold} is specified, \code{alpha} should not be.}
\item{lodcolumn}{If \code{format="onepheno"}, this indicates the LOD
score column to focus on. This should be a single number between 1
and the number of LOD columns in the object input.}
\item{pvalues}{If TRUE, include columns with genome-scan-adjusted
p-values in the results. This requires that \code{perms} be
provided.}
\item{ci.function}{For formats \code{"tabByCol"} and \code{"tabByChr"},
indicates the function to use to get approximate confidence intervals
for QTL location.}
\item{\dots}{For formats \code{"tabByCol"} and \code{"tabByChr"},
additional arguments are passed to the function indicated by
\code{ci.function} (for example, \code{drop} for \code{\link{lodint}}
or \code{prob} for \code{\link{bayesint}}, or \code{expandtomarkers}
for either).}
}
\details{
This function is used to report loci deemed interesting from a one-QTL
genome scan (by \code{\link{scanone}}).
For \code{format="onepheno"}, we focus on a single LOD score column,
indicated by \code{lodcolumn}. The single largest LOD score peak on
each chromosome is extracted. If \code{threshold} is specified, only
those peaks with LOD meeting the threshold will be
returned. If \code{perms} and \code{alpha} are specified, a threshold
is calculated based on the permutation results in \code{perms} for the
significance level \code{alpha}. If neither \code{threshold} nor
\code{alpha} are specified, the peak on each chromosome is returned.
Again note that with this format, only the LOD score column indicated
by \code{lodcolumn} is considered in deciding which chromosomes to
return, but the LOD scores from other columns, at the position with
maximum LOD score in the \code{lodcolumn} column, are also returned.
For \code{format="allpheno"}, we consider all LOD score columns, and
pull out the position, on each chromosome, showing the largest LOD
score. The output thus may contain multiple rows for a chromosome.
Here \code{threshold} may be a vector of LOD score thresholds, one for
each LOD score column, in which case only those positions for which a
LOD score column exceeded its threshold are given. If
\code{threshold} is a single number, it is applied to all of the LOD
score columns. If \code{alpha} is specified, it must be a single
significance level, applied for all LOD score columns, and again
\code{perms} must be specified, and these are used to calculate the
LOD score threshold for the significance level \code{alpha}.
For \code{format="allpeaks"}, the output will contain, for each
chromosome, the maximum LOD score for each LOD score column, at the
position at which it achieved its maximum. Thus, the output will
contain no more than one row per chromosome, but will contain the
position and maximum LOD score for each of the LOD score columns.
The arguments \code{threshold} and \code{alpha} may be specified as
for the \code{"allpheno"} format. The results for a chromosome are
returned if at least one of the LOD score columns exceeded its
threshold.
For \code{format="tabByCol"}, there will be a separate table for each
LOD score column, with a single peak per chromosome. Included are
columns indicating chromosome, peak position, lower and upper limits
of the confidence interval calculated via \code{\link{lodint}} or
\code{\link{bayesint}}, and lod score.
The output for \code{format="tabByChr"}, is similar to that of
\code{format="tabByCol"}, but with results organized by chromosome
rather than by LOD score column.
If \code{pvalues=TRUE}, and \code{perms} is specified,
genome-scan-adjusted p-values are calculated for each LOD score
column, and there are additional columns in the output containing
these p-values.
In the case that X-chromosome specific permutations were performed
(with \code{perm.Xsp=TRUE} in \code{\link{scanone}}), autosome-
and X-chromosome specific thresholds and p-values are calculated by
the method in Broman et al. (2006).
}
\value{
An object of class \code{summary.scanone}, to be printed by
\code{print.summary.scanone}.
}
\author{Karl W Broman, \email{broman@wisc.edu}}
\examples{
\dontshow{set.seed(70377314)}
data(fake.bc)
\dontshow{fake.bc <- subset(fake.bc, chr=c(2,5,6))}
fake.bc <- calc.genoprob(fake.bc, step=5)
# genome scan by Haley-Knott regression
out <- scanone(fake.bc, method="hk")
# permutation tests
\dontrun{operm <- scanone(fake.bc, method="hk", n.perm=1000)
}\dontshow{operm <- scanone(fake.bc, method="hk", n.perm=100)
}
# peaks for all chromosomes
summary(out)
# results with LOD >= 3
summary(out, threshold=3)
# the same, but also showing the p-values
summary(out, threshold=3, perms=operm, pvalues=TRUE)
# results with LOD meeting the 0.05 threshold from the permutation results
summary(out, perms=operm, alpha=0.05)
# the same, also showing the p-values
summary(out, perms=operm, alpha=0.05, pvalues=TRUE)
##### summary with multiple phenotype results
out2 <- scanone(fake.bc, pheno.col=1:2, method="hk")
# permutations
\dontrun{operm2 <- scanone(fake.bc, pheno.col=1:2, method="hk", n.perm=1000)
}\dontshow{operm2 <- scanone(fake.bc, pheno.col=1:2, method="hk", n.perm=100)
}
# results with LOD >= 2 for the 1st phenotype and >= 1 for the 2nd phenotype
# using format="allpheno"
summary(out2, thr=c(2, 1), format="allpheno")
# The same with format="allpeaks"
summary(out2, thr=c(2, 1), format="allpeaks")
# The same with p-values
summary(out2, thr=c(2, 1), format="allpeaks", perms=operm2, pvalues=TRUE)
# results with LOD meeting the 0.05 significance level by the permutations
# using format="allpheno"
summary(out2, format="allpheno", perms=operm2, alpha=0.05)
# The same with p-values
summary(out2, format="allpheno", perms=operm2, alpha=0.05, pvalues=TRUE)
# The same with format="allpeaks"
summary(out2, format="allpeaks", perms=operm2, alpha=0.05, pvalues=TRUE)
# format="tabByCol"
summary(out2, format="tabByCol", perms=operm2, alpha=0.05, pvalues=TRUE)
# format="tabByChr", but using bayes intervals
summary(out2, format="tabByChr", perms=operm2, alpha=0.05, pvalues=TRUE,
ci.function="bayesint")
# format="tabByChr", but using 99% bayes intervals
summary(out2, format="tabByChr", perms=operm2, alpha=0.05, pvalues=TRUE,
ci.function="bayesint", prob=0.99)
}
\seealso{ \code{\link{scanone}}, \code{\link{plot.scanone}},
\code{\link{max.scanone}}, \code{\link{subset.scanone}},
\code{\link{c.scanone}}, \code{\link{summary.scanoneperm}}
\code{\link{c.scanoneperm}} }
\references{
Broman, K. W., Sen, \if{latex}{\out{\'S}}\if{html}{\out{Ś}}\if{text}{S}, Owens, S. E., Manichaikul, A.,
Southard-Smith, E. M. and Churchill G. A. (2006) The X chromosome in
quantitative trait locus mapping. \emph{Genetics}, \bold{174}, 2151--2158.
}
\keyword{print}
|