Scatter plot

The scatter plot, as its name implies, displays a scatter plot where the coordinates are read either from a dataset text file, usually some tsne plot, or are dynamically chosen from the summary plot (see below). To control features of the plot, to the top left there is a burguer menu, where you can set the plot width and height or the sample size, for example. The color and shape attributes allow you to select a dictionary variable to color or shape samples based on the variable values. You can even use a regression method when using the summary plot, to see relations between the variables.

Down to the left there are tools to zoom in/out the plot and move it. You can use these buttons or the mouse to interact with the plot. To zoom with the mouse you need to press the Ctrl key first. The download button allows you to download the plot image. The lasso toggle allows you to select samples and, for example, create a filter or create groups of samples to do analysis. These groups will be also added to the groups tab, where you can create custom variables based on your selection of samples. The filter allows you to apply a global filter and can be edited later from the filter tab. After lassoing you will see a menu with options to edit the samples, add a group and/or use this group to create a filter. If you select to edit the samples a table is displayed where you can remove samples from the selection or rename the group.

Once a group is created you can access menu options like 'Survival Analysis' or 'Summarize' to open new plots for the analysis of the samples selected. From these plots you can also edit the samples by clicking on the edit option from the dictionary variable.

If Survival Analysis is selected a survival plot for the group menu would open on top of the scatter plot.

If Summarize by age group:

To the right, you can see the legend, displaying the categories found for the dictionary terms, with their corresponding colors or shapes. You can edit the colors, and show or hide categories, interacting with this legend.

When using numeric terms to color the samples, continuous mode can be set on the term, in which case a color gradient is used to color the samples.