The software accepts data in comma-separate values (csv) format. Columns should represent features/variables and rows should represent conditions or samples, see Table 1. Data should have a column header indicating feature names (shaded blue in Table 1) and row header indicating condition names (shaded grey in Table 1). The data in ShapoGraphy should be scaled between 0 and 1.0. Dimensional data (i.e. length, width, ..) should be scaled between 0.1 and 1.0 to avoid diminishing of the smallest object. Dimensional data should be scaled in a way that their ratios (w/l) before and after normalization is equal. For example, if the largest object length is l1=100 and width w1=50, then l1'=1 and w1=.5. You can also normalize your data in our app using the Normalise data sub-menu on the right. There, you can select the dimensional variables if applicable.
|Condition||Cell length||Cell width||Number of nuclei||Nucleus length||Nucleus width|
To plot your data using ShapoGraphy you need to first upload data. Currently ShapoGraphy support comma separated files (i.e. csv files, see File format). Add objects to represent map your data to using the shapes on the left menu (e.g. an ellipse for a cell or tumour, a square for a well, or draw your own shape).
The color of an object can be static (fixed for all samples) or dynamic (vary across samples depending on a selected variable value). If you wish to use a static color, you can adjust it from the color picker in the left menu where you can also adjust the object stroke color.
Several features can be mapped to each object as detailed below. The user can show/hide a certain visual element by clicking on the eye icon.
For best experience, run ShapoGraphy using Google Chrome.
You can select variables To sort plotted glyph sets based on the value of a designated variable.
To arrange plotted glyph sets in a scatter plot, you can select variables to position them on the x- and y-axis. This help sorting the glyphs or further group them and effectively add additional dimensions.
The object menu contains a list of the objects you have created
From here, you can manipulate your objects as follows:
The global features menu contain features that will be equally applied to the object across all data points (rows in the data file)
Available global features include:
Object Positions relative to the center of the cluster can also be specified interactively by dragging and dropping the object
An additional option to load your data to our page is selecting the load project option from the file menu
this will open a window where you can select your data.csv file and any previous template.json file you worked on earlier
this will load the dataset and apply the template to it at the same time
Please notice that the template.json file must contain values with similar feature names. If you load a dataset with different feature names then you need to map your features to the object elements again.
You can view the legend specifying the feature mapping to different glyph objects by clicking the legend button in the top menu
You can also save this legend as a SVG file or a PNG image by selecting the save legend option from the file menu
Addition views in ShapoGraphy include:
Accessed by clicking the heatmap button in the top menu.
Shows the dataset values in a table for ease of access
Accessed by clicking the Table icon option in the top menu*.
*The dataset option becomes visible only after loading a data file
The user can zoom in or zoom out on any point on the canvas
This is done by clicking the Zoom in icon and then moving the mouse pointer to the desired point and scrolling the mouse wheel to zoom in or out
To cancel the zooming operation, simply click the Zoom out icon which will revert the canvas to it's original state
Shapography introduces pre-designed templates which the user can use with their own datasets
On the homepage, you can find them in the templates section, click on the template that you like to see and it will be rendered on the canvas before you
At that point, the template will be using a mock dataset for the sake of demonstration, you can load one of our datasets to apply the template to them, or simply use your own
At the moment most templates have their own tutorials which would be indicated at the top of the rendered canvas, you can also access them through the 'Help' tab in the navbar
When a large dataset that contains a minimum of 50 rows or more, the user can use the average per cluster feature
This feature will allow the user to cluster the data using the selected method and show clusters based on the average of the generated clusters values