Comparing flow and imaging for the same question

The same labelled antibody can feed a count or a picture, and those are different claims. The decision this guide supports is which instrument is allowed to carry the sentence you want, and which sentence you should refuse even if the figure looks tidy. The counting path is how a flow cytometer reads a cell. The picture's controls are fluorescence microscopy without the myths.

Both classes of instrument are in the scientific instruments catalogue. The shared biology may sit in the molecular biology pathway. A combined specification goes in a quote request.

Numbers versus places

Flow cytometry measures many cells, one at a time, and reports distributions: a percentage above a gate, a median in a channel, a relationship between two parameters. It is the right tool when the sentence contains "what fraction" or "how bright, across the population". It destroys location. The pulse is a sum.

Imaging keeps location and morphology. It is the right tool when the sentence contains "where" or "what shape", within the resolution of the objective and the thickness of the optical section. It is a poor census. The fields are few, and the eye chooses interesting ones. A claim that needs both a fraction and a place needs both methods, plus a statement of how the cells in the picture relate to the gate in the file. A sorted population plated afterward is one honest link. A random pair of figures with the same antibody name is not.

One question, two sample states

A suspension that runs on a cytometer has been dissociated. Junctions, polarity and tissue architecture are gone. A section or a coverslip culture keeps those and cannot pass a nozzle. If the question is about architecture, do not dissociate and then wish the cytometer had kept it. If the question is about the fraction of cells expressing a marker in a lymph node, a single section is a spatial sample, not that fraction. Design the preparation for the sentence.

Fluorophore choice can be shared and the controls cannot. On the cytometer you need single stains at least as bright as the sample, unstained cells for autofluorescence, and a fluorescence-minus-one gate where the boundary is close. On the microscope you need the negative specimen at the same exposure, the right filter cube, and a scale bar from calibrated pixel size. Compensation is not a brightness slider on the cytometer, and gamma is not a quantitative correction on the image. Keep those adjustments in their own notebooks.

A short decision procedure

Write the sentence. Underline the word that is a number and the word that is a place. If only a number is load-bearing, plan flow, including viability and doublet gates so the number is not dead cells and pairs. If only a place is load-bearing, plan imaging, including the objective aperture and whether widefield haze would fake the place. If both are load-bearing, decide which method is primary and what the other is allowed to illustrate.

Then write the refusal. You will not call colocalisation an interaction. You will not call twenty positive cells in one field a percentage. You will not call a cytometer double-positive a nuclear translocation. These refusals prevent the figure legend from outrunning the method. Optical background for the imaging half is in the MicroscopyU fluorescence section and the Evident fluorescence primer.

When one method's failure tempts you to switch

If the cytometer shows no separation, imaging will not invent a percentage. Check threshold, compensation and viability first. If the microscope is dim, a cytometer can test whether the cells are labelled at all, and the light path still needs its own checks. If the methods disagree on "positive", compare the gates and the exposures before you cite either. A fluorescence-minus-one boundary and a matched imaging negative often show that one of the two positives was a control failure.

Do not pool a flow percentage and an imaging score into one significance test. They are not replicates of the same measurement. Report them as two claims with two control sets.

One question that needs both sentences

You want to know whether a treatment increases the fraction of cells that hold a transcription factor, and whether that factor sits in the nucleus. The fraction is a cytometer sentence: fix and permeabilise by the clone sheet, exclude dead cells with a dye applied before fixation, remove doublets, and set the boundary with a fluorescence-minus-one control. The nuclear location is a microscope sentence on a matched coverslip: the same clone if the sheet allows it, a nuclear counterstain, a scale bar from calibrated pixel size, and a negative slide at the same exposure. If the cytometer fraction rises and the images show only cytoplasmic stain, do not average the disagreement. Check that permeabilisation reached the nucleus and that the focus was not a plane of background. If the images look nuclear and the cytometer shows no shift, check compensation, autofluorescence and whether the parent gate threw the cells away. The two methods are allowed to correct each other. They are not allowed to be blended into one number.

Safety and research limits

Choosing an instrument does not choose a biosafety level. Unfixed human material on either platform is an institutional decision, with the WHO Laboratory biosafety manual as a reference. This guide does not approve it. Neither method, used as research, is a diagnosis. Counting standards in cytometry are part of the discussion at the International Society for Advancement of Cytometry.

What to send with an enquiry

State the sentence, the sample form (suspension, adherent cells, or tissue), the fluorophores, and whether you need a percentage, a location, or a sort followed by a picture. Ask for the cytometer's lasers and filters and the microscope's aperture and camera pixel size as separate lines. Use the scientific instruments catalogue and the quote request. Ask whether a quotation is possible. One instrument name cannot answer both sentences.