Description Usage Arguments Details Value Editing Bins Author(s) Examples
scanGel
plots the raw data from a set of fsa files
in a slab-gel format, optionally including a set of bins. The GUI
interface provides options for users to interactively add and delete
bins.
1 |
pt |
A peak table, as produced by |
bin.lines |
The bin boundaries to use. Can be a two column matrix
or data.frame, with the first column being the lower boundary and the
second the upper boundary. Omit if you want to define bins completely
manually, otherwise use the output of |
Electropherograms from multiple samples are collated into a single gel-like image, such as would have been produced on a slab-gel style instrument. If bin lines are provided, they will be overlaid on the gel image. The user may then add or delete bins with the mouse.
This function does not return anything, but is used to launch
an interactive graphical viewer displaying the data in a peak table.
However, see below for a discussion of the object editedBins
,
which is placed directly in the global environment to 'return' the
modified bins to the user for further use.
Once launched, the viewer displays each sample in a column, and the
y-axis indicates the size in base pairs of individual fragments. (If
you see an empty box, click the refresh
button at the top to
replot the image.) Samples aren't labelled, by design, as when
evaluating scoring you should not be considering the identify of
individuals, but rather the quality of the data (you knew that,
right?).
The height of the electropherogram peak for each fragment is captured by both the color and width of the plotted band. Line width increases with the log of the peak height, so fatter bands are higher peaks. Further, color is used to indicate absolute ranges:
cyan. Note that some very low peaks are effectively invisible, although they may have been used to define bins. This causes some bins to appear empty. Such peaks are filtered out before data analysis, so not a reason for concern.
red
green
black
dark blue
Later in the analysis you will be able to set a threshold to eliminate short peaks. Most studies reject peaks less than 100 or 150 RFU, corresponding to the cyan, red and possibly green bands on the image.
If you have provided the bin.lines
argument, your bins will be
displayed as well. The lower boundary of a bin is indicated with a
dashed orange line, and the upper boundary with a dotted purple line.
In addition, a purple bar is placed in the margin indicating the size
of the bin.
The initial view starts at 50-62 bp. You can page up and down through the data with the toolbar buttons, as well as zoom the view in and out.
Why would you want to edit the bins by hand? In short, because no algorithm will provide a perfect binning of your data. Rather than trust to the objectivity of an arbitrary algorithm, I choose to remove bins that look to have inaccurately grouped fragments.
More rarely, I may split a single large bin into two, or otherwise manipulate the boundaries to better reflect the fragment pattern on the gel. Comparing the results of my hand-edited binning vs the automated binning, there is rarely a noticeable difference. But if you're obsessive about your data, you may appreciate having the ability to exclude the worst, least defined groups of fragments.
In the example, scroll up to base pairs 82-85. If you feel comfortable with the algorithmic binning of this section, no problem. However, you may prefer to exclude this section, as the boundaries between successive bins are not clear.
If you click the Delete Bin
button on the tool bar, you will be
able to remove one bin from the view. Simply click anywhere in the area
of the bin. The entire bin will be shaded out. As you do this, the
modified bin list will be written to your global environment with the
name editedBins
. If you click the Refresh
button, you
will see the revised data set with the deleted bin removed.
Add Bin
works similarly, except that you mark your new bin by
dragging the mouse across the gel to define the upper and lower
boundaries with a temporary rectangle.
Note: there is no undo! However, the variable you passed in to
scanGel
as the bin.lines
argument remains unchanged. So
you can easily return to the initial bins if you don't like the changes
you've made. For this reason, you should *not* use editedBins
as
the argument for scanGel
!
On the other hand, there is no automatic saving of your new bins
either. So if you want to keep you modified bins for future use, be
sure to save editedBins
to a file for safe-keeping.
Tyler Smith
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | ## Not run:
mybins <- fsaRGbin(oxyPT)
scanGel(oxyPT, mybins)
## You may have to click 'Refresh' to see the gel!
## Any changes made to the bins in the GUI will be stored in the
## variable editedBins:
dim(mybins)
dim(editedBins)
## Delete some bins
dim(mybins) ## unchanged
dim(editedBins) ## deleted bins are removed
## If you want to store your changes, make sure to save editedBins to a
## csv file or .Rdata object!!
## End(Not run)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.