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
|
\name{plot.scantwo}
\alias{plot.scantwo}
\title{Plot LOD scores for a two-dimensional genome scan}
\description{
Plot the results of a two-dimensional, two-QTL genome scan.
}
\usage{
\method{plot}{scantwo}(x, chr, incl.markers=FALSE, zlim, lodcolumn=1,
lower = c("full", "add", "cond-int", "cond-add", "int"),
upper = c("int", "cond-add", "cond-int", "add", "full"),
nodiag=TRUE, contours=FALSE, main, zscale=TRUE, point.at.max=FALSE,
col.scheme = c("redblue","cm","gray","heat","terrain","topo"),
gamma=0.6, allow.neg=FALSE, alternate.chrid=FALSE, \dots)
}
\arguments{
\item{x}{An object of class \code{"scantwo"}, as output by
\code{\link[qtl]{scantwo}}.}
\item{chr}{Optional vector specifying which chromosomes to plot.}
\item{incl.markers}{If FALSE, plot LOD scores on an evenly
spaced grid (not including the results at the markers).}
\item{zlim}{A vector of length 2 (optional), indicating the z limits
for the lower-right and upper-left triangles, respectively. If one
number is given, the same limits are used for both triangles. If
\code{zlim} is missing, the maximum limits are used for each.}
\item{lodcolumn}{If the scantwo results contain LOD scores for
multiple phenotypes, this argument indicates which to use in the
plot.}
\item{lower}{Indicates which LOD scores should be plotted in the lower
triangle. See the details below.}
\item{upper}{Indicates which LOD scores should be plotted in the upper
triangle. See the details below.}
\item{nodiag}{If TRUE, suppress the plot of the scanone output
(which is normally along the diagonal.)}
\item{contours}{If TRUE, add a contour to the plot at 1.5-LOD below
its maximum, using a call to \code{\link[graphics]{contour}}. If a
numeric vector, contours are drawn at these values below the maximum
LOD.}
\item{main}{An optional title for the plot.}
\item{zscale}{If TRUE, a color scale is plotted at the right.}
\item{point.at.max}{If TRUE, plot an X at the maximum LOD.}
\item{col.scheme}{Name of color pallet.}
\item{gamma}{Parameter affecting range of colors when
\code{col.scheme="gray"} or \code{="redblue"}.}
\item{allow.neg}{If TRUE, allow the plot of negative LOD scores; in
this case, the z-limits are symmetric about 0. This option is
chiefly to allow a plot of difference between LOD scores from
different methods, calculated via \code{\link[qtl]{-.scantwo}}.}
\item{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.}
\item{\dots}{Ignored at this point.}
}
\value{None.}
\details{
Uses \code{\link[graphics]{image}} to plot a grid of LOD scores. The
particular LOD scores plotted in the upper-left and lower-right
triangles are selected via \code{upper} and \code{lower},
respectively. By default, the upper-left triangle contains the
epistasis LOD
scores (\code{"int"}), and the lower-right triangle contains the LOD
scores for the full model (\code{"full"}).
The diagonal contains either all zeros or the main effects LOD scores
(from \code{\link[qtl]{scanone}}).
The \code{\link[qtl]{scantwo}} function calculates, for each pair of
putative QTLs, \eqn{(q_1,q_2)}{(q1,q2)}, the likelihood undering the
null model \eqn{L_0}{L0}, the likelihood under each of the single-QTL
models, \eqn{L(q_1)}{L(q1)} and \eqn{L(q_2)}{L(q2)}, the likelihood
under an additive QTL model, \eqn{L_a(q_1,q_2)}{La(q1,q2)}, and the
likelihood under a full QTL model (including QTL-QTL interaction),
\eqn{L_f(q_1,q_2)}{Lf(q1,q2)}.
The five possible LOD scores that may be plotted are the following.
The epistasis LOD scores (\code{"int"}) are \eqn{LOD_i = \log_{10}
L_f(q_1,q_2) - \log_{10} L_a(q_1,q_2)}{LODi = log10 Lf(q1,q2) -
log10 La(q1,q2)}.
The full LOD scores (\code{"full"}) are
\eqn{LOD_f = \log_{10} L_f(q_1,q_2) - \log_{10} L_0}{%
LODj = log10 Lf(q1,q2) - log10 L0}.
The additive LOD scores (\code{"add"}) are
\eqn{LOD_a = \log_{10} L_a(q_1,q_2) - \log_{10} L_0}{%
LODa = log10 La(q1,q2) - log10 L0}.
In addition, we may calculate, for each pair of
chromosomes, the difference between the full LOD score and the
maximum single-QTL LOD scores for that pair of chromosomes
(\code{"cond-int"}).
Finally, we may calculate, for each pair of
chromosomes, the difference between the additive LOD score and the
maximum single-QTL LOD scores for that pair of chromosomes
(\code{"cond-add"}).
If a color scale is plotted (\code{zscale=TRUE}), the axis on the
left indicates the scale for the upper-left triangle,
while the axis on the right indicates the scale for the
lower-right triangle. Note that the axis labels can get screwed up
if you change the size of the figure window; you'll need to redo the
plot.
}
\section{Output of addpair}{
\bold{Note} that, for output from \code{\link[qtl]{addpair}} in which the
new loci are indicated explicitly in the formula, the summary provided
by \code{plot.scantwo} is somewhat special. In particular, the
\code{lower} and \code{upper} arguments are ignored.
In the case that the formula used in \code{\link[qtl]{addpair}} was
not symmetric in the two new QTL, the x-axis in the plot corresponds
to the first of the new QTL and the y-axis corresponds to the second
of the new QTL.
}
\examples{
data(hyper)
\dontshow{hyper <- subset(hyper, chr=c(1,4,6,15))}
hyper <- calc.genoprob(hyper, step=5)
\dontshow{hyper <- calc.genoprob(hyper)}
# 2-d scan by EM and by Haley-Knott regression
out2.em <- scantwo(hyper, method="em")
out2.hk <- scantwo(hyper, method="hk")
# plot epistasis and full LOD scores
plot(out2.em)
# plot cond-int in upper triangle and full in lower triangle
# for chromosomes 1, 4, 6, 15
plot(out2.em, upper="cond-int", chr=c(1,4,6,15))
# plot cond-add in upper triangle and add in lower triangle
# for chromosomes 1, 4
plot(out2.em, upper="cond-add", lower="add", chr=c(1,4))
# plot the differences between the LOD scores from Haley-Knott
# regression and the EM algorithm
plot(out2.hk - out2.em, allow.neg=TRUE)
}
\seealso{ \code{\link[qtl]{scantwo}},
\code{\link[qtl]{summary.scantwo}}, \code{\link[qtl]{plot.scanone}},
\code{\link[qtl]{-.scantwo}} }
\author{Hao Wu; Karl W Broman,
\email{kbroman@biostat.wisc.edu}; Brian Yandell }
\keyword{hplot}
|
