Up to package v.0.0.9, there were three plotting functions for the output of QTLscan, namely plotQTL, plotLinearQTL and plotLinearQTL_list.
Since release 0.1.0, the functionality of all three functions has been combined into a single general plotting function, named plotQTL.
The plot layout is now specified by a new argument layout, allowing the user to plot results for single chromosomes separately, or together either adjacently or in a grid layout.
Results from multiple analyses can be overlaid. Previously, it was possible to call the function plotQTL multiple times and overlay subsequent plots using the argument overlay = TRUE.
This approach is no longer supported. Instead, if multiple results are to be overlaid, they can be provided as a list of QTLscan or singleMarkerRegression outputs. Note however that this
is only possible using the default layout. If significance thresholds are
present, the default behaviour is to rescale LOD values so that multiple plots can be combined with overlapping signficance thresholds. This rescaling behaviour can also be
disabled (by setting rescale = FALSE). Note that not all arguments may be appropriate for all layouts.
plotQTL(
LOD_data,
layout = "l",
inter_chm_gap = 5,
ylimits = NULL,
sig.unit = "LOD",
plot_type = "lines",
colour = c("black", "red", "dodgerblue", "sienna4"),
add_xaxis = TRUE,
add_rug = TRUE,
add_thresh = TRUE,
override_thresh = NULL,
thresh.lty = 3,
thresh.lwd = 2,
thresh.col = "darkred",
return_plotData = FALSE,
show_thresh_CI = FALSE,
use_LG_names = TRUE,
axis_label.cex = 1,
custom_LG_names = NULL,
LGdiv.col = "gray42",
ylab.at = 2.5,
highlight_positions = NULL,
mainTitle = FALSE,
rescale = TRUE,
...
)plotLinearQTL(
LOD_data,
layout = "l",
inter_chm_gap = 5,
ylimits = NULL,
sig.unit = "LOD",
plot_type = "lines",
colour = c("black", "red", "dodgerblue", "sienna4"),
add_xaxis = TRUE,
add_rug = TRUE,
add_thresh = TRUE,
override_thresh = NULL,
thresh.lty = 3,
thresh.lwd = 2,
thresh.col = "darkred",
return_plotData = FALSE,
show_thresh_CI = FALSE,
use_LG_names = TRUE,
axis_label.cex = 1,
custom_LG_names = NULL,
LGdiv.col = "gray42",
ylab.at = 2.5,
highlight_positions = NULL,
mainTitle = FALSE,
rescale = TRUE,
...
)
plotLinearQTL_list(
LOD_data,
layout = "l",
inter_chm_gap = 5,
ylimits = NULL,
sig.unit = "LOD",
plot_type = "lines",
colour = c("black", "red", "dodgerblue", "sienna4"),
add_xaxis = TRUE,
add_rug = TRUE,
add_thresh = TRUE,
override_thresh = NULL,
thresh.lty = 3,
thresh.lwd = 2,
thresh.col = "darkred",
return_plotData = FALSE,
show_thresh_CI = FALSE,
use_LG_names = TRUE,
axis_label.cex = 1,
custom_LG_names = NULL,
LGdiv.col = "gray42",
ylab.at = 2.5,
highlight_positions = NULL,
mainTitle = FALSE,
rescale = TRUE,
...
)
The plot data, if return_plotData = TRUE. Otherwise NULL. Output is returned invisibly
Output of QTLscan function. If you wish to overlay multiple genome-wide QTLscan outputs, then first compile these into a single list and
pass this to LOD_data, for example LOD_data = list(qtl_res1, qtl_res2). If this is passed as a named list and add_legend = TRUE, these names will be used
in the legend as well.
There are three possible plot layouts - single chromosome plots ("s"), genome-wide plots arranged adjacently in a linear fashion ("l") which is
also the default, and genome-wide plots arranged in a grid ("g"), i.e. a grid of single chromosome plots. In the latter case, a suitable grid dimension will be determined
based on the number of linkage groups detected in LOD_data. If you wish to overlay results from multiple multi-chromosome analyses, use the default layout.
The gap size (in units of cM) between successive chromosomes when layout = "l". By default a gap of 5 cM is used. Normally the user should not need to change this.
Use to specify ylimits of plot region, though by default NULL in which case a suitable plot region is automatically used.
Label to use on the y-axis for significance units, by default assumed to be LOD score.
Plots can be either in line drawings ("lines", default) or scatter plot format ("points").
Vector of colours to be used in the plotting. A default set of 4 colours is provided, the first of which is used when results from a single QTL scan are to be plotted.
Should an x-axis be drawn? If multiple QTL analyses are performed on different traits, specifying this to be FALSE
and using par(mar=c(0,4.1,4.1,2.1)) allows subsequent plots to be neatly stacked.
Logical, by default TRUE - should original marker points be added to plot?
Logical, by default TRUE - should a significance threshold be added to plot?
By default NULL. Can be used to specify a (numeric) value for the significance threshold, overriding any stored in LOD_data. If you wish to override
thresholds for multiple analyses (so, when LOD_data is a list of QTL outputs), can also provide a vector of numeric values here.
Gives user control over the line type of the significance threshold to be drawn. Default threshold lty is 3.
Gives user control over the line width of the significance threshold to be drawn. Default threshold lwd is 2.
Gives user control over the line colour of the significance threshold to be drawn. Default threshold colour is dark red. If plotting multiple analyses with rescale = FALSE,
it can be useful to provide the same colours to this argument as to colour, so that LOD profiles can be linked to their respective LOD thresholds.
Logical, by default FALSE. If TRUE, then the x and y coordinates of the plot data are returned when layout = "l",
which can be useful for subsequent plot manipulations and overlays. For other layouts, no plot data is returned.
Logical, by default FALSE. Should confidence interval bounds around LOD threshold be shown if available? If LOD_data is a list from multiple analyses,
this option is ignored to prevent plot becoming too cluttered.
Logical, by default TRUE. Should original character LG names (the names of list LOD_data) be used as axis labels? If FALSE, numbering is used instead.
Argument to adjust the size of the axis labels. Can be useful if there are many linkage groups to plot
Option to specify a vector that contains custom linkage group names. By default NULL. See previous argument use_LG_names, which is the usual manner
for controlling x-axis labels.
Colour of dividing lines between linkage groups when layout = "l", by default grey.
Distance from the y-axis to place label (by default at 2.5 points)
Option to include a (list of) positions to highlight (e.g. peak QTL positions). Each list element should be a 2-column data.frame with columns giving
the linkage group numbers (numeric) and the corresponding cM positions (numeric) to highlight. If LOD_data is the result of a single genome-wide scan, it is also possible to just directly provide the
2-column data.frame (again, with column 1 containing linkage group numbers and column 2 containing corresponding cM positions).
If LOD_data has been provided as a list of multiple analyses, you may wish to highlight different positions from each analysis. Then highlight_positions should also be a list
of the same length and in the same order as LOD_data. Each data.frame of positions will be coloured in the same colour as the LOD output.
If no position is to be highlighted for some analyses, add the corresponding list element as NULL. For example, if you wish to highlight positions for analyses 1 and 3 in a 3-analysis overlay,
then use something like highlight_positions = list(data.frame(lg = 1, cM = 50),NULL,data.frame(lg=c(2,3),cM=c(11,99))).
The default setting is NULL, meaning no positions are highlighted.
Option to supply vector of plot titles if layout = "s" or layout = "g". Argument ignored if using the default layout. Single character vector also allowed and will be recycled.
For no plot titles, leave as default, i.e. FALSE
If results from multiple analyses are to be overlaid and different significance thresholds are detected, then by default plots will be rescaled so that threshold lines overlap.
This behaviour can be disabled by setting rescale = FALSE.
Arguments passed to plot, and lines or points as appropriate (see argument plot_type).
if (FALSE) {
data("qtl_LODs.4x")
plotQTL(LOD_data = qtl_LODs.4x,layout = "l")
}
Run the code above in your browser using DataLab