Classification

Classification in Partek Flow

The Garnett¹ automated cell type classification algorithm has been wrapped into Partek Flow as the Classification task. As with the original Garnett tool, Classification in Partek Flow works on single-cell data, along with a cell type definition (marker) file, and trains a regression-based classifier. Once a classifier is obtained and published, it can be applied to classify future datasets from similar tissues. To improve the user experience, both the marker file (.txt) and the classifier file (.rds) have been implemented as library files in Partek Flow.

Garnett classifiers in Partek Flow

Partek Flow hosts a selection of pre-trained classifiers as Managed classifiers. The list of available classifiers can be found here². If a managed classifier exists for your data type, we recommend that you try it. In addition to the Managed classifiers, the Project classifiers trained on the same dataset prior to your classification, may also be used to classify cell type. Project classifiers can be promoted to Managed classifiers if users publish them.

Classify cell type task in Partek Flow

To classify cell type using a pre-trained classifier:

Select any non-normalized single cell data node. Filtered counts, which contains the raw count, is used here. Next, choose the Classify cell type task in the Classification section (Figure 1)

Figure 1. Selecting the Classify cell type task.

If using the Managed classifiers tool for the first time in Partek Flow, you will be asked to create a new classifier file (Figure 2a). Users may select either the Download Garnett classifier that matches the species and tissue type with their dataset from a Partek Flow maintained list or the Import Garnett classifier that is trained out of Partek Flow (Figure 2b). Next, push the Create button to create the classifier file. Once the correct classifier file has been created, select Finish to start running the task (Figure 2d).

Figure 2. Create Garnett classifier from Managed classifiers and run Classify cell type task with the selected classifier.

If users would like to classify cell type with the classifier obtained from the same project, ‘Project classifiers’ needs to be selected from the dropdown list. Next, use ‘Select data node’ to choose the classifier before clicking the Finish button (Figure 3).

Figure 3. Run Classify cell type task with Project classifier.

Regardless of which type of classifier was applied, Partek Flow will output a new data node named Classify result (Figure 4). These outputs of cell type annotation are exactly the same as Garnett¹. Downstream analysis tasks such as normalization and PCA can be performed on the Classify result data node.

Figure 4. Classify cell type task outputs a new data node.

Train classifier in Partek Flow

To train a classifier with a list of biomarkers with your own dataset:

Select any non-normalized single cell data node, Filtered counts is used here. Next, choose the Train classifier task in the Classification section (Figure 1)

Similar to previously described steps, first time users will be asked to create the Marker file. Partek Flow does not currently host any marker files, however, users may add them as library files. Marker files should be a .txt file with the marker information in correct format. The same example in the Garnett tutorial of a simple valid Marker file is provided here (Figure 5). For more details on how to construct a Marker file, please refer to Garnett tutorial³. Next, click the Create button and Partek Flow will then save the file with the name that users provide for future use (Figure 6a).

Figure 5. Example of marker file (.txt) from Garnett tutorial.

After the proper Marker file has been chosen, click the Finish button to start running the task as default (Figure 6b).

Figure 6. Train Classifier task in Flow.

Train classifier task report in Partek Flow

Once the task has finished, click the Classifier data node and choose the Task report in the Task results section, or prompt it to be one of the Managed classifiers by clicking the Publish classifier task in the Classification section (Figure 7).

Figure 7. Train classifier task report in Task results section.

Users will find two parts in the task report: the marker evaluation plot (Figure 8) and the classification gene table (Figure 9). The marker evaluation plot provides some key information about whether the chosen markers are optimal. Ambiguity scores are calculated for each of the markers which indicates how many cells receive ambiguous labels when this marker is included. The classification gene table may give a hint to which genes are chosen as the relevant genes for distinguishing between different cell types.

Figure 8. Example of biomarker evaluation plot in task report.

Figure 9. Example of classification gene table in task report.

Other adjustable parameters in this task include:

Number of Unknown: this tells Garnett how many outgroup cells it should compare against; the default is 500. For a dataset with fewer cells, the number should be smaller.

References

Additional Assistance

If you need additional assistance, please visit our support page to submit a help ticket or find phone numbers for regional support.

Partek Flow Documentation

Page tree