The format should follow the starting frequency and the downstream assay — not the brand you used last time.
The two formats answer different questions. Column-free negative selection asks: can I remove everything that is not my target, without touching the target? Column-based positive selection asks: can I capture my target specifically enough that what comes off the column is pure?
This is the single strongest predictor. If the target is already 30–60% of the suspension — mouse splenic T cells, human PBMC T cells, splenic B cells — negative selection removes the minority and leaves a clean majority: typical purity lands at 90–97% with viability above 90%, in roughly half an hour, and the cells never see an antibody against their own surface marker.
Once the target drops below about 10%, negative selection runs out of room: the depletion cocktail has to remove more than 90% of the sample, and every residual contaminant is a large relative error. This is the regime for column-based positive selection. A CD34⁺ fraction in PBMC (0.5–2%) is the textbook case; a rare population at 0.15% is the stress test, where a single column typically reaches 40–60% and a second column on a fresh cartridge is what pushes purity past 85%.
Positive selection puts a nanobead on the target cell. With beads around 50 nm that occupy only 20–30% of the antigen epitopes, flow-cytometry panels still work and no specific antibody clone is required — but if your readout is the target antigen itself, or if you plan to stain with a competing clone, plan the panel around the bead.
Negative selection is the safer default when the cell state is the readout: chemotaxis, NET formation, TCR signalling and any assay sensitive to cross-linking.
Target above ~10% and state-sensitive readout → column-free negative selection. Target below ~10% or purity is the bottleneck → column-based positive selection, with a second column if the first lands under 60%. When both matter, enrich first and sort second: beads reduce the number of cells going onto the sorter, which is usually the expensive step.
Tell us your cell type, sample source and downstream assay — we will suggest an isolation route, a culture condition and the matching catalogue numbers.