In this tutorial, we demonstrate how to:
The tutorial is based on the work published by Venteicher and co-workers, on isocitrate dehydrogenase-mutant gliomas. Single cells from tumor biopsies were processed by flow cytometry and the libraries were prepared by Smart-seq2 protocol. The tutorial data set consists of eight expression matrix files, one per patient sample. The tumors were categorized as either astrocytoma or oligodendroglioma glioma subtype by histology. The matrix files contain gene expression values normalized by the following transformation log2[(TPM/10)+1].
If you need additional assistance, please visit our support page to submit a help ticket or find phone numbers for regional support.
The tutorial data set is available through Partek Flow.
Click your avatar (Figure 1)
Click Settings
On the System information page, the Download tutorial data section includes pre-loaded data sets used by Partek Flow tutorials (Figure 2).
Click Single cell glioma (multi-sample)
The tutorial data set will be downloaded onto your Partek Flow server and a new project, Glioma (multi-sample), will be created. You will be directed to the Data tab of the new project. Because this is a tutorial project, there is no need to click on Import data, as the import is handled automatically (Figure 3).
You can wait a few minutes for the download to complete, or check the download progress by selecting Queue then View queued tasks... to view the Queue (Figure 4).
Once the download completes, the sample table will appear in the Data tab, with one row per sample (Figure 5).
The sample table is pre-populated with two sample attributes: # Cells and Subtype. Sample attributes can be added and edited manually by clicking Manage in the Sample attributes menu on the left. If a new attribute is added, click Assign values to assign samples to different groups. Alternatively, you can use the Assign values from a file option to assign sample attributes using a tab-delimited text file. For more information about sample attributes, see here.
For this tutorial, we do not need to edit or change any sample attributes.
With samples imported and annotated, we can begin analysis.
Click Analyses to switch to the Analyses tab
For now, the Analyses tab has only a single node, Single cell counts. As you perform the analysis, additional nodes representing tasks and new data will be created, forming a visual representation of your analysis pipeline.
Click on the Single cell counts node
A context-sensitive menu will appear on the right-hand side of the pipeline (Figure 9). This menu includes tasks that can be performed on the selected counts data node.
An important step in analyzing single cell RNA-Seq data is to filter out low-quality cells. A few examples of low-quality cells are doublets, cells damaged during cell isolation, or cells with too few counts to be analyzed.
Expand the QA/QC section of the task menu
Click on Single cell QA/QC (Figure 6)
A task node, Single cell QA/QC, is produced. Initially, the node will be semi-transparent to indicate that it has been queued, but not completed. A progress bar will appear on the Single cell QA/QC task node to indicate that the task is running.
Click the Single cell QA/QC node once it finishes running
Click Task report on the task menu (Figure 7)
The Single cell QA/QC report opens in a new data viewer session. There are interactive violin plots showing the most commonly used quality metrics for each cell from all samples combined (Figure 8). For this data set, there are two relevant plots: the total count per cell and the number of detected genes per cell. Each point on the plots is a cell and the violins illustrate the distribution of values for the y-axis metric. Typically, there is a third plot showing the percentage of mitochondrial counts per cell, but mitochondrial transcripts were not included in the data set by the study authors, so this plot is not informative for this data set.
Remove the % mitochondrial counts and the extra text box in the bottom right by clicking Remove plot in the top right corner of each plot (Figure 8).
The plots are highly customizable and can be used to explore the quality of cells in different samples.
Click on Single cell counts in the Get Data icon on the left (Figure 9)
Click and drag the Sample name attribute onto the Counts plot and drop it onto the X-axis
Repeat this for the Detected genes plot
The cells are now separated into different samples along the x-axis (Figure 10)
Hold Control and left-click to select both plots
Open the Style icon on the left under Configure
Under Color, use the slider to reduce the Opacity
Open the Axis icon on the left
Adjust the X-rotation on the plots to 90
Note how both plots were modified at the same time.
Cells can be selected by setting thresholds using the Select & Filter tool. Here, we will select cells based on the total count
Open Select & Filter under Tools on the left
Under Criteria, Click Pin histogram to see the distribution of counts
Set the Counts thresholds to 8000 and 20500
Selected cells will be in blue and deselected cells will be dimmed (Figure 11).
Because this data set was already filtered by the study authors to include only high-quality cells, this count filter is sufficient.
Click Apply observation filter
Click the Single cell counts data node in the pipeline preview (Figure 12)
Click Select
A new task, Filter counts, is added to the Analyses tab. This task produces a new Filter counts data node (Figure 13).
Click on the Glioma (multi-sample) project name at the top to go back to the Analyses tab
Your browser may warn you that any unsaved changes to the data viewer session will be lost. Ignore this message and proceed to the Analyses tab
Most tasks can be queued up on data nodes that have not yet been generated, so you can wait for filtering step to complete, or proceed to the next section.
A common task in bulk and single-cell RNA-Seq analysis is to filter the data to include only informative genes. Because there is no gold standard for what makes a gene informative or not, ideal gene filtering criteria depends on your experimental design and research question. Thus, Partek Flow has a wide variety of flexible filtering options.
Click the Filter counts node produced by the Filter counts task
Click Filtering in the task menu
Click Filter features (Figure 14)
There are four categories of filter available - noise reduction, statistics based, feature metadata, and feature list (Figure 15).
The noise reduction filter allows you to exclude genes considered background noise based on a variety of criteria. The statistics based filter is useful for focusing on a certain number or percentile of genes based on a variety of metrics, such as variance. The feature list filter allows you to filter your data set to include or exclude particular genes.
We will use a noise reduction filter to exclude genes that are not expressed by any cell in the data set but were included in the matrix file.
Click the Noise reduction filter checkbox
Set the Noise reduction filter to Exclude features where value <= 0 in 99% of cells using the drop-down menus and text boxes (Figure 16)
Click Finish to apply the filter
This produces a Filtered counts data node. This will be the starting point for the next stage of analysis - identifying cell types in the data using the interactive t-SNE plot.
We are omitting normalization in this tutorial because the data has already been normalized.
The tutorial data set is taken from a published study and has already been normalized using TPM (Transcripts per million), which normalizes for the length of feature and total reads, and transformed as log2(TPM/10+1). This normalization and transformation scheme can be performed in Partek Flow, along with other commonly used RNA-Seq data normalization methods.
For more information on normalizing data in Partek Flow, please see the Normalization section of the user manual.
If you need additional assistance, please visit our support page to submit a help ticket or find phone numbers for regional support.
Click under Filter to include the selected cells
Differential expression analysis can be used to compare cell types. Here, we will compare glioma and oligodendrocyte cells to identify genes differentially regulated in glioma cells from the oligodendroglioma subtype. Glioma cells in oligodendroglioma are thought to originate from oligodendrocytes, thus directly comparing the two cell types will identify genes that distinguish them.
To analyze only the oligodendroglioma subtype, we can filter the samples.
Click the Filtered counts data node
Expand Filtering in the task menu
Click Filter cells (Figure 1)
The filter lets us include or exclude samples based on sample ID and attribute.
Set the filter to Include samples where Subtype is Oligodendroglioma
Click AND
Set the second filter to exclude Cell type (multi-sample) is Microglia
Click Finish to apply the filter (Figure 2)
A Filtered counts data node will be created with only cells that are from oligodendroglioma samples (Figure 3).
Click the new Filtered counts data node
Click Statistics > Differential analysis in the task menu
Click GSA
The configuration options (Figure 4) includes sample and cell-level attributes. Here, we want to compare different cell types so we will include Cell type (multi-sample).
Click Cell type (multi-sample)
Click Next
Next, we will set up a comparison between glioma and oligodendrocyte cells.
Click Glioma in the top panel
Click Oligodendrocytes in the bottom panel
Click Add comparison (Figure 5)
This will set up fold calculations with glioma as the numerator and oligodendrocytes as the denominator.
Click Finish to run the GSA
A green GSA data node will be generated containing the results of the GSA.
Double-click the green GSA data node to open the GSA report
Because of the large number of cells and large differences between cell types, the p-values and FDR step up values are very low for highly significant genes. We can use the volcano plot to preview the effect of applying different significance thresholds.
Open the Style icon on the left, change Size point size to 6
Open the Axes icon on the left and change the Y-axis to FDR step up (Glioma vs Oligodendrocytes)
Open the Statistics icon and change the Significance of X threshold to -10 and 10 and the Y threshold to 0.001
Open the Select & Filter icon, set the Fold change thresholds to -10 and 10
Note these changes in the icon settings and volcano plot below (Figure 6).
We can now recreate these conditions in the GSA report filter.
Click GSA report tab in your web browser to return to the GSA report
Click FDR step up
Set the FDR step up filter to Less than or equal to 0.001
Press Enter
Click Fold change
Set the Fold change filter to From -10 to 10
Press Enter
The filter should include 291 genes.
To visualize the results, we can generate a hierarchical clustering heatmap.
Click the Filtered feature list produced by the Differential analysis filter task
Click Exploratory analysis in the task menu
Click Hierarchical clustering/heatmap
Using the hierarchical clustering options we can choose to include only cells from certain samples. We can also choose the order of cells on the heatmap instead of clustering. Here, we will include only glioma cells and order the samples by sample name (Figure 7).
Make sure Cluster is unchecked for Cell order
Click Filter cells under Filtering and set the filter to include Cell type (multi-sample) is Glioma
Choose Sample name from the Cell order drop-down menu in the Assign order section
Click Finish
Double click the green Hierarchical clustering node to open the heatmap
The heatmap differences may be hard to distinguish at first; the range from red to blue with a white midpoint is set very wide because of a few outlier cells. We can adjust the range to make more subtle differences visible. We can also adjust the color.
Set the Range toggle Min to -1.5
Set the Range toggle Max to 1.5
The heatmap now shows clear patterns of red and blue.
Click Axis titles and deselect the Row labels and Column labels of the panel to hide sample and feature names, respectively.
Select Sample name from the Annotations drop-down menu
Cells are now labeled with their sample name. Interestingly, samples show characteristic patterns of expression for these genes (Figure 8).
Click Glioma (multi-sample) to return to the Analyses tab.
We can use gene set enrichment to further characterize the differences between glioma and oligodendrocyte cells.
Click the Filtered feature list node
Click Biological interpretation in the task menu
Click Gene set enrichment
Change Database to Gene set database and click Finish to continue with the most recent gene set (Figure 9)
A Gene set enrichment node will be added to the pipeline .
Double-click the Gene set enrichment task node to open the task report
Top GO terms in the enrichment report include "ensheathment of neurons" and "axon ensheathment" (Figure 10), which corresponds well with the role of oligodendrocytes in creating the myelin sheath that supports and protect axons in the central nervous system.
If you need additional assistance, please visit our support page to submit a help ticket or find phone numbers for regional support.
Click to view the Volcano plot
In Select & Filter, click to remove the P-value (Glioma vs Oligodendrocytes) selection rule. From the drop-down list, add FDR step up (Glioma vs Oligodendrocytes) as a selection rule and set the maximum to 0.001
Click to apply the filter and generate a Filtered Feature list node
t-SNE (t-distributed stochastic neighbor embedding) is a visualization method commonly used to analyze single-cell RNA-Seq data. Each cell is shown as a point on the plot and each cell is positioned so that it is close to cells with similar overall gene expression. When working with multiple samples, a t-SNE plot can be drawn for each sample or all samples can be combined into a single plot. Viewing samples individually is the default in Partek Flow because sample to sample variation and outlier samples can obscure cell type differences if all samples are plotted together. However, as you will see in this tutorial, in some data sets, cell type differences can be visualized even when samples are combined.
Using the t-SNE plot, cells can be classified based on clustering results and differences in expression of key marker genes.
Prior to performing t-SNE, it is a good idea to reduce the dimensionality of the data using principal components analysis (PCA).
Click the Filtered counts data node after the Filter features task
Select PCA from the Exploratory analysis section of the task menu (Figure 1)
Click Finish to run PCA with default settings (Figure 2)
Note, the default settings include the Split by sample checkbox being selected. This means that the dimensionality reduction will be performed on each sample separately.
PCA task and data nodes will be generated.
Click the PCA data node
Select t-SNE from the Exploratory analysis section of the task menu (Figure 3)
Click Finish from the t-SNE dialog to run t-SNE with the default settings (Figure 4)
Because the upstream PCA task was performed separately for each sample, the t-SNE task will also be performed separately for each sample. t-SNE task and data nodes will be generated (Figure 5).
Once the t-SNE task has completed, we can view the t-SNE plots
Click the t-SNE node
Click Task report from the task menu or double click the t-SNE node
The t-SNE will open in a new data viewer session. The t-SNE plot for the first sample in the data set, MGH36 (Figure 6), will open on the canvas. Please note that the appearance of the t-SNE plot may differ each time it is drawn so your t-SNE plots may look different than those shown in this tutorial. However, the cell-to-cell relationships indicated will be the same.
The t-SNE plot is in 3D by default. To change the default, click your avatar in the top right > Settings > My Preferences and edit your graphics preferences and change the default scatter plot format from 3D to 2D.
You can rotate the 3D plot by left-clicking and dragging your mouse. You can zoom in and out using your mouse wheel. The 2D t-SNE is also calculated and you can switch between the 2D and 3D plots on the canvas. We will do this later on in the tutorial.
Each sample has its own plot. We can switch between samples.
Open the Axes icon on the left under Configure (Figure 7)
Navigate to Misc
Select the 
icon below the Sample name to go to the next sample
The t-SNE plot has switched to show the next sample, MGH42 (Figure 7).
The goal of this analysis is to compare malignant cells from two different glioma subtypes, astrocytoma and oligodendroglioma. To do this, we need to identify the malignant cells we want to include and which cells are the normal cells we want to exclude.
The t-SNE plot in Partek Flow offers several options for identifying, selecting, and classifying cells. In this tutorial, we will use the expression of known marker genes to identify cell types.
To visualize the expression of a marker gene, we can color cells on the t-SNE plot by their expression level.
Select any of the count data nodes from Get data on the left (Single cell counts, or any of the Filtered counts, Figure 8)
Search for the BCAN gene
Click and drag the BCAN gene onto the plot and drop it over the Green (feature) option
The cells will be colored from black to green based on their expression level of BCAN, with cells expressing higher levels more green (Figure 9). BCAN is highly expressed in glioma cells.
In Partek Flow, we can color cells by more than one gene. We will now add a second glioma marker gene, GPM6A.
Select any of the count data nodes from the Data card on the left (Single cell counts, or any of the Filtered counts)
Search for the GPM6A gene
Click and drag the GPM6A gene onto the plot and drop it over the Red (feature) option
Cells expressing GPM6A are now colored red and cells expressing BCAN are colored green. Cells expressing both genes are colored yellow, while cells expressing neither are colored black (Figure 10).
Numerical expression levels for each gene can be viewed for individual cells.
Switch to pointer mode by clicking 
in the top right corner of the plot
Select a cell by pointing and clicking
The expression level for that cell is displayed on the legend for each gene. Expression values can also be viewed by mousing over a cell (Figure 11).
Deselect the cell by clicking on any blank space on the plot
Now that cells are colored by the expression of two glioma cell markers, we can classify any cell that expresses these genes as glioma cells. Because t-SNE groups cells that are similar across the high-dimensional gene expression data, we will consider cells that form a group where the majority of cells express BCAN and/or GPM6A as the same cell type, even if they do not express either marker gene.
Switch to lasso mode by clicking 
in the top right of the plot
Draw the lasso around the cluster of green, red, and yellow cells and click the circle to close the lasso (Figure 12)
Selected cells are shown in bold and unselected cells are dimmed. The number of selected cells is indicated in the figure legend. The cells are plotted on the color scale depending on their relative expression levels of the two marker genes (Figure 13)
Click Classify selection in the Classify icon under Tools
A dialog to give the classification a name will appear.
Name the classification Glioma
Click Save (Figure 14)
Once cells have been classified, the classification is added to Classify. The number of cells belonging to the classification is listed. In MGH42, there are 460 glioma cells (Figure 15).
Classifications made on the t-SNE plot are retained as a draft as part of the data viewer session. In this tutorial, we will classify malignant cells for each sample before we save and apply the classifications, but if necessary, you can save the data viewer session by clicking the 
Save icon on the left to retain all of the formatting and draft classifications. The data viewer session will be stored under the Data viewer tab and can be re-opened to continue making classifications at a later time.
Switch to pointer mode by clicking in the top right corner of the plot
Deselect the cells by clicking on any blank space on the plot
Open Axes and navigate to Sample under Misc
Select the icon below the sample name to go to the next sample, MGH45
Rotate the 3D t-SNE plot to get a better view of cells from the green, red, and yellow cluster
Switch to lasso mode by selecting in the top right corner of the plot
Draw the lasso around the cluster of colored cells and click the circle to close the lasso (Figure 16)
Select Classify selection in the Classify icon
Type Glioma or select Glioma from the drop-down list (Figure 17)
Click Save
Repeat these steps for each of the 6 remaining samples. Remember to go back to the first sample (MGH36) to classify the glioma cells in that samples too.
There should be 5,322 glioma cells in total across all 8 samples.
The classification name can be edited or deleted (Figure 18).
With the malignant cells in every sample classified, it is time to save the classifications.
Click Apply classifications in the Classify icon
Name the classification attribute Cell type (sample level)
Click Run (Figure 19)
The new attribute is stored in the Data tab and is available to any node in the project.
Click on the Glioma (multi-sample) project name at the top to go back to the Analyses tab
Your browser may warn you that any unsaved changes to the data viewer session will be lost. Ignore this message and proceed to the Analyses tab
For some data sets, cell types can be distinguished when all samples can be visualized together on one t-SNE plot. We will use a t-SNE plot of all samples to classify glioma, microglia, and oligodendrocyte cell types.
Click on the Glioma (multi-sample) project name at the top to go back to the Analyses tab
Click the Filtered counts data node after the Filter features task
Click PCA in the Exploratory analysis section of the task menu
Uncheck the Split by sample checkbox (Figure 22)
Click Finish
The PCA task will run as a new green layer.
Click the new PCA data node
Select t-SNE from the Exploratory analysis section of the task menu
Click Finish to run the t-SNE task with default settings
The t-SNE task will be added to the green layer (Figure 23). Layers are created in Partek Flow when the same task is run on the same data node.
Once the task has completed, we can view the plot.
Double-click the green t-SNE data node to open the t-SNE scatter plot
Click and drag the 2D scatter plot icon onto the canvas and replace the 3D scatter plot (Figure 24)
Search for and select green t-SNE data node (Figure 25)
In the Style icon, choose Sample name from the Color by drop-down list under Color
Viewing the 2D t-SNE plot, while most cells cluster by sample, there are a few clusters with cells from multiple samples (Figure 26).
Using marker genes, BCAN (glioma), CD14 (microglia), and MAG (oligodendrocytes), we can assess whether these multi-sample clusters belong to our known cell types.
Select any of the count data nodes from the Data card on the left (Single cell counts, or any of the Filtered counts)
Search for the BCAN gene
Click and drag the BCAN gene onto the plot and drop it over the Green (feature) option
Search for the CD14 gene
Click and drag the CD14 gene onto the plot and drop it over the Red (feature) option
Search for the MAG gene
Click and drag the MAG gene onto the plot and drop it over the Blue (feature) option
After coloring by these marker genes, three cell populations are clearly visible (Figure 27).
The red cells are CD14 positive, indicating that they are the microglia from every sample.
Switch to lasso mode by clicking the icon in the top right of the plot
Draw the lasso around the cluster of red cells and click the circle to close the lasso (Figure 28)
Open the Classify tool and click Classify selection
Name the classification Microglia
Click Save
The blue cells are MAG positive, indicating that they are the oligodendrocytes from every sample.
Switch to pointer mode by clicking in the top right corner of the plot
Deselect the cells by clicking on any blank space on the plot
Switch to lasso mode again by clicking the icon in the top right of the plot
Draw the lasso around the cluster of blue cells and click the circle to close the lasso
Open the Classify tool and click Classify selection
Name the classification Oligodendrocytes
Click Save
Finally, we will classify the BCAN expressing cells on the plot as glioma cells from every sample.
Switch to pointer mode by clicking in the top right corner of the plot
Deselect the cells by clicking on any blank space on the plot
Switch to lasso mode again by clicking the icon in the top right of the plot
Draw the lasso around the cluster of green cells and click the circle to close the lasso
Open the Classify tool and click Classify selection
Name the classification Glioma
Click Save
Switch to pointer mode by clicking in the top right corner of the plot
Deselect the cells by clicking on any blank space on the plot
The number of cells classified as microglia, oligodendrocytes, and glioma are shown in Classify (Figure 29)
Click Apply classifications in the Classify icon (Figure 30)
Name the classification attribute Cell type (multi-sample) (Figure 31)
Click Run
The new attribute is now available for downstream analysis.
Click on the Glioma (multi-sample) project name at the top to go back to the Analyses tab
Your browser may warn you that any unsaved changes to the data viewer session will be lost. Ignore this message and proceed to the Analyses tab
If you need additional assistance, please visit our support page to submit a help ticket or find phone numbers for regional support.