Having trouble analysing your high-dimensional flow cytometry data? Getting an accurate understanding of large or complex information requires automated tools rather than manual gating on bivariate dot plots. Tercen provides advanced high-dimensional data analysis using the latest algorithms. This ready-to-use analysis platform needs no knowledge of coding, therefore it removes the barrier between laboratory researchers and the world of bioinformatics. In this article we show you how to connect your FlowJo workspace to Tercen.
Some things you need to check before you start.
You will need a full license of FlowJo installed on your computer.
Make sure you have an open connection to the internet.
No time to read?
Watch a video of how it's done:
Install the Tercen plugin.
Step 1: Download the Tercen plugin folder and extract its contents.
Download the Tercen plugin (tercen.zip) from FlowJo Exchange.
Open the downloaded folder and extract the files. There will be two files.
(N.B. Your Tercen plugin file may have different version numbers).
This PDF contains installation instructions.
Step 2: Add the Tercen plugin file to your FlowJo plugin folder
To install the Tercen plugin, move or copy the Tercen plugin file (e.g. Tercen_v1.0.0.jar) to the FlowJo plugins folder.
The location of the plugin folder differs depending on your operation system.
Refer to the FlowJo Plugins Documentation to set up or locate your plugins folder.
Step 3: Restart FlowJo
Close and re-open FlowJo to activate the plugin.
The Tercen plugin will now be listed in the Plugins drop-down menu.
Upload from FlowJo to Tercen
Step 1: Select a population.
N.B. This guide does not attempt to show you how to use FlowJo software and we recommend you consult the FlowJo website for user manuals and tutorials.
We recommend you QC and clean your data with FlowJo tools before uploading to Tercen. Removing abnormal data points, debris, doublets and dead cells in advance will speed up the uploading process and help you get a more accurate analysis.
After performing your preferred cleaning and gating strategy in FlowJo select a population of your interest by clicking on it with your mouse.
Step2: Apply the Tercen plugin to the selected population.
From the Workspace tab press the Plugin button and choose Tercen from the dropdown menu.
A popup with Instructions on how to use the Tercen plugin will appear.
Press OK to continue.
A Tercen connector will now be created for your chosen population.
Step3: Copy the Tercen connector to more populations.
Tercen can upload multiple populations or raw data-files simultaneously. The connector must be applied to each selection individually.
The connector can be applied to other populations individually by using drag and drop with your mouse or to all samples by using the copy analysis to group function in FlowJo.
Step 4: Upload the connected data to Tercen
To upload data, double-click on any of the Tercen connectors.
A popup window will appear to allow you to refine your selected data..
The upper window contains all files or populations which currently have a connector.
The plugin will default to uploading all connected samples.
Use your mouse to select or deselect populations as required.
The lower window contains the parameters (channels) which will be grouped for analysis.
Compensated parameters are automatically selected by default.
Use your mouse to select or deselect channels as required.
When you are happy with your selection click OK to initiate the upload.
An upload progress bar will appear in a popup window.
A note on using multiple files:
Uploading multiple .fcs files into the same analysis is only possible when they have the same channel markers. Tercen will give an error message if it sees files which do not have the same structure.
Making a Tercen Account.
If you are new to Tercen the plugin will create an account for you.
You will be able to use this Tercen account from any internet browser.
Tercen will default to your FlowJo software license details if they are available.
Follow the instructions on screen to set a password and choose a username.
Step 5: Select an analysis workflow template
When the data upload is successful Tercen will open a page in your browser.
Tercen will create a default name for your workflow. You can change it here.
Next there are some pre-made template workflows for you to choose from. These templates are designed for immunophenotyping. They explore the characteristics of a high-dimensional flow cytometry dataset using clustering and dimensionality reduction tools.
There are two templates available.
The FlowJo Cluster Explorer #1 - Basic
This workflow contains steps for a standard clustering analysis using self-organizing maps (FlowSOM) and has visualizations of the clustering results in a heatmap format.
The FlowJo Cluster Explorer #2 - Advanced
In addition to the clustering steps described above, this workflow contains steps which perform dimensionality reduction, using tSNE and UMAP. It includes visualizations of their results. It also presents an additional heatmap, created using MEMscore values, that allows the evaluation of the characteristics of each flowSOM cluster.
When you have selected an analysis template press OK.
Tercen creates a unique project name using your FlowJo Workspace name. Projects are not overwritten if data is re-uploaded, a new project is created instead. This means you will be able to return to your original data analysis even if you make later modifications to gating or population selections in FlowJo and choose to re-upload these.
Step 6: Run the analysis and get the result
The workflow canvas will open showing the steps of the template.
Your data is already loaded into the table step.
In the global task bar select run all
Tercen will now run all of the analysis steps in the template.
When the workflow is finished each step will have a green dot to say the analysis was successful.
The bottom steps of the workflow contain the visualizations.
If you picked the Basic workflow, double-click on FloSOM Heatmap. It shows you the marker expression level and pattern in each cluster as well as the size of the cluster within the sample.
For the Advanced workflow we recommend you open the Fit-SNE per marker and UMAP per marker to see an overview of marker expression level and pattern. Then open the FlowSOM MEM Heatmap to see a scoring of the clusters based on marker enrichment. Then right-click on the FlowSom Heatmap & MST step and select Run to further explore the characteristics of the clusters.
Understanding the pre-made workflows
Workflows contain computation and visualization steps. The former are used to perform analytical processes and the latter to explore the results of the computation steps through an interactive display.
Common to Basic and Advanced
Data labelling: Tagging each event with a random number between 0 and 100.
Downsampling: Reducing the amount of events for the analysis (optional, used to speed up the analysis).
FlowSOM: Clustering flow cytometry data using Self-Organizing Maps (SOMs).
Cluster re-order: Reorganizing the computed FlowSOM clusters in hierarchical order.
MEM score: Calculating enrichment scores to assess the enrichment of each marker within cell populations.
Flt-SNE: a statistical technique for dimensionality reduction
UMAP: a statistical technique for dimensionality reduction
NB: Some algorithms (for example UMAP) can take a while to process large amounts of data. Depending on how large your data set is, you might need to leave the system running for a while.
When all boxes have a green dot you can explore your results. The following data steps are the result viewers and the short explanation of what you would expect from those views. Double-click a data step to see the visualization it has made.
FlowSOM heatmap: FlowSOM generated clusters visualized in a heatmap format colored by the channels (markers) expression level. The clusters are organized in hierarchical order.
FlowSOM heatmap and MST : In addition to the heatmap, FlowSOM output is visualized in the minimum spanning tree (MST) that depicts population abundances and channel (marker) expressions.
FlowSOM MEM heatmap: FlowSOM generated clusters visualized in a heatmap format using the values calculated by MEM score and colored by the score of the channels (markers).The clusters are organized in hierarchical order.
Flt-SNE FlowSOM color: Visualization of Flt-SNE results in a 2D plot format per sample, colored by FlowSOM cluster.
Flt-SNE per marker: Visualization of Flt-SNE results in a 2D plot format per channel (marker) and sample.
UMAP FlowSOM color: Visualization of UMAP results in a 2D plot format per sample, colored by FlowSOM cluster.
UMAP per marker: Visualization of UMAP results in a 2D plot format per channel (marker) and sample.
Note: To view FlowSOM MST, select the Operator view tab on the left side of the screen.
Adjusting the Downsampling algorithm
We have added a Downsampling step to the workflows. This allows you to get a quick overview of your data and check visualizations.
The method being used is called Uniform Downsampling and, by default it reduces the sample size to 1% of data.
To adjust the sample size follow these steps.
Note: You will need to reset the Downsampling step if it has already run (i.e. has a green dot). Right-click with your mouse and select Reset from the menu before continuing below.
Double-Click on the Downsampling Data Step (or right-click and select Open) to open it.
Click on the Downsample Control Filter
Change the Filter Settings parameter to the percentage of your data (1-100%) you wish to be included in the calculations.
Save your changes by pressing the orange save icon.
Return to your workflow.
Click the Play icon to run all steps.
The Tercen plug-in will stay up-to-date.
Before uploading data Tercen checks that you have the latest version of the plugin.
If we have released a new feature since you installed the plugin then Tercen will download a new one automatically and prompt you with a popup message to restart your FlowJo.