This is an interactive visualization, which is constantly under development (thus the dev). In general to use the function follow,
Loading the experiment is as easy as double clicking the RD.experiment.Rdata1
within the folder. Performing this action with automatically set the correct working directory.
If you would like to circumvent this, it is essential you set the correct working directory, then load the RD.experiment.Rdata
. Below is an example, after opening the R console, or an R terminal.
# Load in the package
require(procPharm)
# Set the working directory (this is where the experiment is located)
# If you have downloaded the experiment from the paper, check the download folder.
setwd("Y:/Lee Leavitt/Presentations/Paper/Figures/TTA A2 Dose Response/190531.39.m.m3.p2 ttaa2 potassium dose response comparison/")
# Load the experiment.
load("./RD.190531.39.m.m3.p2.Rdata")
For an overview of the composition of the experiment data, go here.
Now that both the software and the experiment data is loaded invoking the tcd()
function. This function has flexible inputs and many features to use during visualization as well as analysis tasks.
Most consoles are interactive. To view what has populated the current work space us the ls()
function. The output of this function will show the experiment name.
ls()
prior to using tcd familiarize yourself with the controls
?tcd
To enter the experiment for visualization, (the gif shows the initial view entering the visualization),
tcd(RD.190531.39.m.m3.p2)
up: Moves through current selected list of cell
down: Moves down through selected list of cells
1-9, -, =: Collect cells into groups 1-12
o Order traces in single view based on c.dat, bin, scp. For information on these data frames please go here
F7: Populate the groups with the cell_types
shiftp: Select group of cells to populate the groups
s: Stack all cells from the selected group
shifts: Sample stack selected group. Usefule if selected group has more than 50 cells.
Additionally the cells can be viewed using,
v View the cells from the selected group
shifti Change the image for the multi-view and the single-view images. Only one image can be selected here.
i select the image to display next to the stacked traces. Can be multiple
p The traces are a combination of points and line. To toggle the points press this button.
Visualizing traces has many more options.
t select the type of trace to view
h select the color of the trace
shiftd Change the separation of the traces < 1 closer together > 1 further apart
shifto Sort the traces on a continuous selected variable.
r Rename a selected group to what you would like.
shiftv Select the values to appear on the right side of the trace.
u Underline toggle for both stacked traces and single trace view
l Select the window regions to display. Any or all can be observed.
w To change the line width of the traces pressing this button will allow you to enter in a line width. Default value is 2.
In the example below, * A new group is chosen using shiftp * The baseline corrected trace is selected t * The color of the trace is selected h * The underline is removed u * The traces are squished closer together shiftd * More values are added to the right of trace using shiftv
One of the main functionalities of this software is to create groups. Combined with the custom statistics below, rapidly identifying cells pertaining to a specific response type is easy.
Pressing any of these buttons will places cells into groups 1 through 12.
1,2,3,4,5,6,7,8,9,0,-,=
Pressing any of these keys will remove neurons from groups 1 through 12.
!, @, #, $, %, ^, &, *, (, ), _, +
r Rename: Groups are generically named g.names#
. These names are uninformative. So, to changes these names pressing r will first ask which group to rename. Then the console will ask for the new name.
m Following the custom statistics and box-plot localization, cells are placed in g.names12
or simply the 12th group regardless of the current composition of the group. So to move the new group into a better location pressing m will move the cells to the newly specified group.
Importantly, on a rare occassion tcd
will crash. This function has an BACKUP
object created very often during the time spent in tcd. To resume from where you left off simply perform this,
tcd(RD.experiment, BACKUP)
To leace tcd press q. A question will prompt asking "Would you like yo save your groups".
Sorting traces based on non obvious characteristics is important. To capture these characteristics two functions have been created to make these statistics. These functions are encapsulated in the F1 and F2 key.
This function allows you to select various window regions to create a ratio. During the experiments we apply compounds which elicit amplification, block, or direct effects. Rapidly sorting the traces based on these effects is paramount for a rapid analysis.
The function additionally has a "boxPlotSelector" functionality. This means. one can use a box plot to select cells based on the new statistic.
The user has the option to save the statistic. This appends the newly created statistic to the end of the scp data frame with the user input name. Once this statistic is saved the user has the option of sorting cells based on this stat.
Before starting this function, the cells currently selected will be the cells sent into this function. Meaning the output of this function is completely dependent on the input to this function.
The workflow of this functionality is,
1. The **one** option will return all cells above the click
2. The **two** option will return cell between the clicks. This helps to localize cell based on the statistic.
Once complete the selected cells are now sorted based on this statistic and placed within the 12th group. To Select these cells press shiftp, and then select g.names12 or the twelfth group regardless of the name.
Moving groups is also easy now. To move the group press m. The first question is, which group to move. The second question is the new group to move it to. This will over write the group.
This statistic follows closely what the F1 custom statistic does. The difference is only two windows are compare and a min max normalization is performed. This statistic creates a range between -1 and 1. Placing everythin within this range prevents missing out on extreme outliers. **This function does not require clicking the stop locator
at the top left.
(Red + Blue) / (Red - Blue)
The general work flow of this example is,
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.