Build and review gates

Start from a known sample

Open a sample from its workspace so the workbench has the correct event data and saved analysis. Opening the workbench without a sample starts a synthetic demo. Check the sample name before interpreting any plot.

Review compensation before defining fluorescence gates. Set the axes and transforms for the parameters you need to inspect. A display transform changes how values appear on the plot; it is a separate choice from compensation.

Create a population hierarchy

1. Select the parent population in the gating tree. The next gate selects events within that parent.
2. Choose the plot axes and a gate tool appropriate to the population, then draw the boundary on the plot.
3. Give the gate a descriptive name and check its event count and frequency. Confirm which parent is used as the denominator.
4. Select the new population to inspect it on another pair of channels. Continue building the hierarchy as required by your experiment.

Review before propagating

Check the boundaries on controls and representative samples from each condition. Review sparse populations and events near gate edges. If you use Auto-gate, inspect its proposed boundary against the underlying events.

Use the propagation controls to apply a strategy to other samples only after checking the parameter mapping. Revisit the resulting plots: the same boundary may behave differently when a population shifts or sample quality changes.

Save and verify

Save the analysis and check its saved state before leaving. Keep population names, parent relationships, compensation, and transforms with the experiment. When comparing reported frequencies, use the same reference population across samples.

Check the denominator explicitly

A population frequency is only interpretable with its reference population. For example, 100 selected events within a parent containing 1,000 events gives 10% of that parent. If the file contains 10,000 events, the same selection is 1% of all events. These are different measurements.

Record the parent path with exported frequencies. A matching gate name in two samples does not establish that both gates have the same parent or select equivalent events.

Review after a gate change

1. Inspect the selected gate and its event count.
2. Walk through each descendant population and check whether its count or frequency changed.
3. Review the same hierarchy on controls and representative samples from each condition.
4. Revisit downstream statistics and figures that depended on the previous population selection.
5. Save the reviewed state and record why the boundary changed.

Troubleshoot an empty or unexpected population

Start at the root and find the first parent where events disappear. Check the active sample, axes, gate coordinates, compensation, and transforms at that point. Moving a descendant gate cannot recover events already removed by an upstream gate.

If propagation behaves differently across samples, compare the parameter mapping and sample distribution. Keep any sample-specific adjustment explicit in the experiment record.