What is spillover spreading, and how does it limit a multicolor panel?

Why

Compensation removes the average spillover but not its noise. The leftover error grows with the square root of the spilled signal. Nguyen and colleagues derived a value for each pair of detectors that does not depend on how bright the reference sample is.

They collected all pairs in a spillover spreading matrix, where each row is a fluorochrome and each column a detector. A large value means that fluorochrome badly widens the negative population in that detector. Panel designers read the matrix in two directions.

Reading down a column shows which detectors receive a lot of spread, and these are poor homes for dim markers. Reading along a row shows which dyes spread widely, and these should go on markers that are not co-expressed with critical dim ones. Spread only hurts when the cells are bright in the spilling dye and are the same cells you read in the receiving detector.

For example, CD4 on BV421 spreading into BV510 matters little for CD4 negative cells but can hide a dim marker on CD4 T cells. The values are specific to one instrument and its settings, and they change after service or a voltage change.