A biomarker positivity annotation is useful for an indication of cell phenotype, state, or functional status which is defined by the binary (+/-) classification of expression of a single biomarker. These may be used as part of determining a cell phenotype annotation, or orthogonally to define cell status across multiple phenotypes. A biomarker positivity annotation should label every cell in a dataset, with the exception of cells that were excluded from analysis due to data quality reasons by Artifact Detection or Cell Quality Control.

For Insight reports that require a biomarker positivity annotation, you can either purchase our annotation services or use the Enable Cloud Platform to generate the annotations yourself. The guidelines below outline the steps you can take on our platform to generate this type of annotation.

Since spatial proteomics is semi-quantitative, it is often necessary to define a custom biomarker expression threshold for each image. The best approach to generating this type of annotation on our platform typically uses the gating app.

You may want to start by generating an automated gating hierarchy, using a template like the one below:

{
	"Ki67":{},
	"notKi67":{}
}

With "Ki67" replaced with the name of your biomarker of interest in your biomarker panel.

This will create a binary gate with the autogating algorithm’s best guess at the appropriate threshold to mark cells as biomarker positive or negative within each specific region. You may wish to inspect the resulting gates at this stage to ensure that the gates were drawn appropriately.

After visual inspection in the gating app, you should export the gates and inspect the annotation in the visualizer by selecting the biomarker of interest and the overlay generated from the gating app export in the channels panel of the visualizer. This step is critical, since noise and imperfections in segmentation can lead to misleading results at the biomarker expression level—visual inspection of the annotation on the image is the most reliable way to ensure cells are being classified correctly. You can create a visualizer story noting on each region which gates have been drawn too generously (too many cells marked positive), or too conservatively (too few marked positive). You may want to select both the positive & negative cell annotations so you can determine whether an object you see not being marked positive has been excluded because of gating, or because it was missed by segmentation (will not appear in the negative annotation).

If you find that the gates need adjustment after exporting, you can return to the gating app and generate a new “Manual” hierarchy choosing the “Copy an existing hierarchy” option and selecting your previously exported hierarchy. This maintains your previous hierarchy in place for reference purposes while creating a new editable hierarchy that you can use for further gating refinements.