How to reproduce a flow cytometry panel from a paper

This guide describes how to rebuild a published flow cytometry panel so that your gates mean the same thing as the authors' gates. Papers often list marker names without fluorophores, clones, or controls, and show a gating figure without the logic that produced it, which yields a panel that looks similar and reports different frequencies. Below are the fields to recover and the failures that follow when they are guessed.

What methods sections usually leave out

Fluorophore assigned to each marker, plus the antibody clone
Marker names alone do not define a panel. Assigning a dim marker to a weak fluorophore, or using a different clone, changes resolution and can move a population across a gate.
Which controls were used for setting gates
Unstained cells, isotype controls, and fluorescence-minus-one controls justify different gate placements. Without knowing which was used, your positive gate is not their positive gate.
Compensation approach and the compensation controls
Single-stain controls on the same cells or on beads, and whether autofluorescence was included, change the spillover matrix and therefore the apparent frequency in multi-color plots.
The full gating hierarchy
A frequency only means something relative to its parent gate. Papers report percentages without always stating whether the denominator was all events, all cells, live cells, or a lineage gate.
Viability staining and how dead cells were handled
Dead cells bind antibodies non-specifically and can fall anywhere in the plots. If dead cells were excluded in the paper but not in your run, frequencies shift.
Number of events collected and the instrument configuration
Rare population frequencies are unstable at low event counts, and detector configuration determines whether a given fluorophore combination is even resolvable.

Key parameters and what they change

Parameter	Typical	Changes together with
Cells per stain	Often around one million per tube for a multi-color panel, adjusted to availability	Staining volume and antibody amount per tube; the total cells needed for all conditions and controls.
Antibody amount per test	Titrated per antibody rather than taken from a fixed dilution	Staining volume, resolution between positive and negative populations, and non-specific background.
Staining volume	A small volume that keeps antibody concentration high enough	Antibody amount for a given concentration. Doubling the volume at the same concentration doubles the antibody needed.
Number of control tubes	Unstained, viability-only, single-stain compensation controls, and one fluorescence-minus-one control per color when gates are ambiguous	Total cells and total antibody required. Panel size drives the control count, which drives everything else.
Events collected per sample	Set by the frequency of the population of interest	Acquisition time per sample and the precision of the reported frequency.

Troubleshooting

Symptom	Likely causes	What to check first
Populations appear diagonal or shifted in two-color plots	Under-compensated or over-compensated spillover, or compensation controls that do not match the sample autofluorescence.	Recollect single-stain controls on the same cell type, confirm each control is at least as bright as the sample, and recalculate the matrix.
A dim marker cannot be separated from the negative population	The marker is paired with a low-brightness fluorophore, antibody amount is too low, or spillover from a bright channel raises background.	Titrate the antibody, move the dim marker to a brighter fluorophore in a less crowded detector, and check the spillover spreading into that channel.
High non-specific staining and a smeared negative population	Dead cells, Fc receptor binding, insufficient washing, or too much antibody.	Add a viability dye and an Fc block, wash more thoroughly, and run an antibody titration to find the plateau.
Reported frequencies differ from the paper for apparently similar gates	A different parent gate, different denominator, or dead cells included in one analysis and excluded in the other.	Reconstruct the full hierarchy from the paper's figure and state the denominator explicitly for every frequency you report.
Rare population frequency jumps between replicate samples	Too few events collected, so the count is dominated by sampling noise.	Increase events collected until the absolute count of the population of interest is large enough to be stable, and report counts alongside percentages.

Worked example: from a methods sentence to executable steps

Single-cell suspensions were stained with antibodies against surface markers, washed, and analyzed on a flow cytometer. Dead cells were excluded and the frequency of marker-positive cells was determined.

Each step is tagged by how clearly the text states it: explicit, partial, inferred, or missing.

1.
Prepare single-cell suspensionsPartial
Stated, but the dissociation method and buffer are not described.
2.
Fluorophore and clone for each antibodyMissing
Check the reagent table; without this the panel is not defined.
3.
Block Fc receptors before stainingInferred
Standard for many cell types but not mentioned in this text.
4.
Stain with the surface marker panelExplicit
Antibody amounts and staining volume are not given.
5.
Exclude dead cellsPartial
Stated, but not whether a viability dye or scatter gating was used.
6.
Wash and resuspend for acquisitionExplicit
7.
Acquire on the cytometer with compensation appliedPartial
Acquisition is stated; instrument, detector configuration, and compensation controls are not.
8.
Gate and report marker-positive frequencyPartial
The gating hierarchy and the denominator for the reported percentage are unclear.

How Vara helps

Upload the paper or paste the methods text, and Vara builds the panel as steps with per-marker parameters, each tagged explicit, partial, inferred, or missing.
Undefined fields such as fluorophore assignments, antibody amounts, and which controls were used appear as reproduction risks instead of being filled with a guess.
Cell number, staining volume, antibody amount, and the number of control tubes are linked by formulas, so adding a color or a sample rescales the whole stain plan.
The experiment is laid out as lanes of sub-experiments with dependencies, so compensation controls stay tied to the stain they belong to, and each acquisition is logged against the protocol version.

Try Vara on your own protocol See how importing works

FAQ

Do I need fluorescence-minus-one controls or are isotype controls enough?+

For multi-color panels, fluorescence-minus-one controls are generally the better basis for gate placement because they show the actual spread from all the other colors in that channel. Isotype controls address non-specific antibody binding, which is a different question, and they are not a substitute.

Why does my gating give a different percentage than the paper?+

Most often the denominator differs. A frequency is only interpretable relative to its parent gate, so check whether the paper reported the population as a fraction of all events, all cells, live cells, or a lineage-gated subset, then rebuild the same hierarchy.

How do I know if my compensation is wrong?+

Look for populations that run diagonally across a two-parameter plot, or a negative population whose median shifts between channels after compensation. Collect single-stain controls on the same cell type as your sample, make sure each is at least as bright as the sample, and recalculate.

How many events should I collect?+

Enough for the rarest population you want to quantify. Precision depends on the absolute number of events in that gate, not the percentage, so for a rare subset the collection target has to rise accordingly. Reporting counts next to percentages makes this visible to readers.