Analyse Data: Workflow¶

Workflows are pipelines that process data, perform calculations on it, make visualizations, and create reports.

Press the New workflow button

Name the workflow. (for example "Example Workflow")

Press Next

Note: Pre-defined workflows can be selected. These are called Templates. Templates are used to standardize pipelines to perform routine analysis or ensure repeatability by following the same analysis steps for each experiment.

The Workflow Canvas will load.

• Save: Will turn orange when there is a change to save.
• Run All: Runs each step in the workflow sequentially until all have completed or failed.
• Reset Unlocks a run-completed workflow so it can be modified.
• Add Step Add an independent Step to the workflow. Use this to start a pipeline.
• Clone Takes a copy of the workflow and recreates it in a different project.
• Size Control Adjust the zoom view of the canvas.

Table Step¶

Start a pipeline by pressing the Add Step button.

Choose the Table Step and press Ok

Select the uploaded Example_Data_File.csv file from Current Project and press Ok.

Note: Data Tables uploaded to the Main Library of Tercen are available to every project on the server. Normally, these are reference data sets for calibration.

The data table is loaded to the workflow canvas. To extend the pipeline, click the step to bring up the Local Toolbar.

• Reset: Resets this step and any dependent steps in the pipeline. Previous steps are not reset.
• Edit: Opens this step to expose its features and settings.
• Rename: Change the step name.
• Delete: Removes the step.
• Duplicate: Makes a copy of the step linked to the same previous data.
• Add: Adds a new step, downstream from this step.
• Status Indicator: Green when that Data Step has run successfully. Red if there is a problem.

Data Step¶

Press the Add button.

Choose the Data Step option and press Ok.

The Data Step will open automatically.

Data Steps are the main engine blocks of analysis in Tercen. Visualizations and Calculations are set up and performed here.

• Control Bar: Features for data manipulation and visualization.
• Factors Panel: Factors from the data file appear at the beginning of this list. As the pipeline is built folders are added containing the new factors that are created by the calculations it makes. Each Data Step has access to the factors of the steps that precede it in the pipeline.
• Crosstab Grid: Factors are dropped onto this grid to make projections. These projections create the base data set that is used to create visualizations or perform calculations.

Make an X-Y Axis projection¶

Drag VALUE out of the Factors panel with your mouse.

The areas where it can be dropped will go green.

The crosstab grid will offer an X-axis and a Y-axis, which project numbers.

It will offer a Row and Column, which are used to form groupings of the data in the X-Y Axes.

We recommend that you always start a projection with your primary measurement on the Y-axis.

Drop VALUE on the Y-axis.

The crosstab builds a Data Cell with all the measurements in the VALUE factor ordered from lowest to highest.

Note: Unlike a spreadsheet data cell, a Tercen data cell can hold multiple data points. This important concept impacts how a calculation or visualization is performed.

Drag CELL_SUBSET to X-Axis to make a simple projection.

Grab the black lines of the crosstab grid and drag them out to make the data more visible.

Hold the CTRL key on your keyboard and scroll with the mouse wheel to zoom in and out.

Press Save and return to the workflow canvas by clicking the breadcrumb.

Click the data Step to bring up the Local toolbar and rename this Data Step to "Simple Projection".

Save the workflow.

Make a Row/Column projection¶

Click the Data Table and add a new Data Step from the Local Toolbar.

When Factors are dropped to columns or Rows, the crosstab splits the data cell into groups of data points according to that factor. These new data cells have their own X-Y Axis.

Multiple Factors can be dropped to Rows and Columns, forming subgroups. A black line will appear on either side of a Row or Column indicating whether a Factor will be dropped before or after the existing one.

Groupings on a Crosstab projection work from Outside to In. For Columns, the topmost factor is grouped first, and then sub-groups in order below it. For Rows, the leftmost factor is grouped first, and then subsequent factors working to the right.

Make the following projection.

VALUE to Y-Axis.
SEX to Column.
AGE to Column (after SEX).
CELL_SUBSET to Row.
PANEL_TYPE to Row (before CELL_SUBSET).

Save the Data Step.

Make a Heatmap¶

The Control Bar has buttons and Zones to control how a projection is visualized.

• Buttons: Apply a pre-defined change to the data such as point size, style, or clear the grid.
• Zones: Light up green and apply a change to the data that is defined by the Factor that is dropped onto it. These are Filters, Colors, Labels, and Error Bar.

Adjust the projection as follows.

SUBJECT_ID to Column (After AGE).
VALUE to Colors.

The crosstab grid now has one data cell per measurement and they are coloured according to their VALUE with low numbers trending towards blue and high numbers trending towards Red.

Press the Style button and change it from Point to Heatmap.

Take a moment to move the black lines and zoom in to examine the data. Try to look for immunological patterns.

Save the Data Step, return to the workflow canvas and rename the data step "Heatmap".

Grab Data Steps and move them on the canvas with your mouse.

Adjust the Colour Palette¶

Tercen has color control options for visualizations made with the crosstab.

Right-click on the data cell to bring up the menu.

Select Palette

Experiment with the choices on offer.

Take a Snapshot¶

Tercen can take a quick snapshot of the visualization in the crosstab grid and export it as a .png image file.

Press the Download button to export the heatmap.

Note: This feature will only export what it sees on screen. If the crosstab is longer than the bottom of your screen those parts will not be imaged. Adjust the black lines to get a better image or follow the tutorial to learn how to use the Plot operator.