Preliminary note and copyright information
The following description originates from our FlowFate publication (currently under review) in STAR Protocols, an open-access protocol journal.
Running the differentiation analysis using FlowFate
Images will follow soon
Within R-studio, launch the FlowFate software app with this command:
flowFate::run_app()A new window will open within RStudio which will launch FlowFate. Click ‘Open in Browser’ to launch the software on the web browser. While the display is enabled by the web browser, all processing happens on your local computer.
Select the data format in which the .fcs files are stored. To upload a merged file containing all controls and target sample data, select ‘Single, merged FCS file’. Click ‘Browse’ to select the file. For uploading individual .fcs files containing control and target sample data separately within one folder, select ‘Folder with individual FCS files’. Click ‘Browse’ and select the folder containing all the files that were acquired within the same experiment. For first time users, it is recommended to study the analysis pipeline, using a demo dataset by clicking ‘Import demo data’.
Using your own .fcs files, click ‘Submit’ to confirm your selection and to view your uploaded data.
Continuing with either your own or demo data. At the top in the menu bar, click ‘Explore’. On the left, select the plot type, the scale (linear or biexponential) and channels to display on the X- and Y-axes. For example, select ‘linear’ scale to plot FSC on the X-axis and SSC on the Y-axis. Since the fluorescent markers are visualized on a logarithmic scale, select ‘biexponential’ for instance for the EGFP-channel (here ‘GRN.B.HLin’) to plot the EGFP+ cells on the X-axis and the MyHC-channel (here ‘RED.R.HLin’) to plot MyHC+ cells on the Y-axis. Note that the plotted data is raw and uncurated, hence it also contains debris.
On the menu bar, click ‘Curate’. Select the channel names from your instrument used for EGFP- and MyHC-detection. Select here in the demo dataset for ‘GFP channel’ ‘GRN.B.HLin’ and for the ‘MyHC channel’ ‘RED.R.HLin’. Importantly, select your control samples corresponding to ‘double-negative control’ and ‘MyHC+ control’ as described in step 6.
‘Start curation’ automatically sets up threshold values for debris, based on the double-negative- and MyHC+ -control samples. Thresholding involves the addition of a pre-defined gate to remove cellular debris. Make sure you have a nice single-cell suspension, as doublet-discrimination is not implemented here. Navigate through the ‘NonDebris gate’ tab to display results from this part of the curation process.
Curation in addition sets a GFP-threshold. Cells above this threshold are EGFP+ and cells below this threshold are EGFP-. This function permits the exclusion of untransfected or autofluorescent cells from the analysis. Click the ‘GFP gate’ tab to display results from that part of the curation process.
Next, GFP-expression level bins are set up. On the menu bar, click ‘Gate’ and then select ‘Add GFP bin’ to add at least a GFP-low bin. Then specifying the desired intensity range of that bin. Click ‘Confirm’. Note that the first value displayed in the GFP-low bin corresponds to the GFP-threshold value and it should remain unaltered.
Optionally, if a gene-dose dependency is intended to be analyzed, you can set two additional bins, GFP-medium and -high. Click on ‘Add GFP bin’ and specify the desired intensity range. For the demo data, specify three GFP bins: GFP-low = GFP channel intensities from 24.1 to 100 RFU; GFP-medium = GFP channel intensities from 101 to 1000 RFU; GFP-high= GFP channel intensities from 1001 to 10000 RFU. Click ‘Confirm’ to set each range. You can reset the GFP-bin assignment by clicking the ‘Reset bins’ button.
After FlowFate has switched to the ‘Split peaks’ tab, click ‘Split’. This important step automatically identifies the MyHC- and MyHC+ subpopulations within the defined GFP-bins.
After clicking ‘Split’, the app automatically opens the ‘Plot’ panel. You can browse data in the window. Target samples ‘03_G12C_DMSO’ to ‘04_G12C_AMG510’ within the demo dataset were considered for the subsequent steps. ‘03_G12C_DMSO’ comprises cells transfected with K-RasG12C treated with 0.2 % DMSO and ‘04_G12C_AMG510’ represents cells transfected with K-RasG12C treated with 3 mM AMG 510.
In the ‘Split peaks’ panel, individual datasets can be selected to display MyHC intensity distribution. Multiple selections are possible by clicking on them to facilitate sample comparisons.
To store your analysis for processing, click ‘Export’ on the menu bar. Click ‘Show population statistics’ to get a table with population data necessary for statistical analysis outside of FlowFate. This table contains three columns, (1) sample name (sample), (2) population name (pop), (3) population count (count). ‘Count’ corresponds to the number of events in the respective population. The nomenclature used to unambiguously identify the different populations is explained in Table 1.
Specify your filename and click ‘Download’ to export the data as a .csv file. In order to calculate the fraction of differentiated cells in the GFP-low bin, employ data from /NonDebris/GFP+/GFP-low/MyHC+ count (MyHC+ cells in GFP-low bin) divided by /NonDebris/GFP+/GFP-low count, which represents the total number of GFP+ cells in the GFP-low bin. Apply this processing rationale to all of your target samples.
| Population name (pop) | Definition |
|---|---|
| root | Total events (raw import data) |
| /NonDebris | Intact cells, thresholded for debris exclusion |
| /NonDebris/GFP+ | Intact cells with GFP intensities above the GFP-threshold; removed untransfected or autofluorescent cells |
| /NonDebris/GFP+/GFP-low (-medium;-high) | Intact cells with GFP intensities inside the GFP-low (-medium; -high) bins |
| /NonDebris/GFP+/GFP-low (-medium;-high) /MyHC+ | Intact cells with GFP intensities inside the GFP-low (-medium; high) bins and differentiated with high MyHC expression |