Proliferation Index Calculator
Calculate proliferation index, CFSE dilution generations, and Ki67 staining interpretation.
How CFSE Dilution Works
CFSE binds cytoplasmic proteins covalently and is distributed roughly equally between daughter cells at each division. Fluorescence therefore halves with every division, producing discrete peaks.
| Generation | Relative MFI | Cells from one original |
|---|---|---|
| 0 | 100% | 1 |
| 1 | 50% | 2 |
| 2 | 25% | 4 |
| 3 | 12.5% | 8 |
| 7–8 | <1% | Detection limit |
Resolution runs out after seven or eight divisions, when signal approaches autofluorescence. Beyond that, peaks merge and counting is unreliable.
Two Different Indices
| Metric | Denominator | Answers |
|---|---|---|
| Division index | All cells | Average divisions per original cell |
| Proliferation index | Responding cells only | Average divisions among those that divided |
| % divided | — | What fraction responded at all |
The distinction matters when interpreting suppression. If a drug reduces the fraction responding but the responders divide normally, the division index falls while the proliferation index does not — indicating fewer cells were recruited rather than slower division. Reporting only one index conflates the two mechanisms.
Why Cell Numbers Must Be Back-Calculated
Each generation contains 2n cells per original precursor, so observed counts must be divided by 2n to recover how many cells started in each generation. Skipping this step massively overweights the later generations, since one original cell that divided three times contributes eight events.
| Confound | Effect |
|---|---|
| Cell death | Later generations lost, understating division |
| Uneven dye partitioning | Peak broadening, merged generations |
| Dye toxicity at high concentration | Suppresses the proliferation being measured |
| Autofluorescence | Sets the practical resolution limit |
CFSE at excessive concentration is itself antiproliferative, so titrating to the lowest workable level is standard. Compared with Ki-67, which marks any cell in cycle at one moment, CFSE gives division history — the two answer different questions.
Worked Examples
Common Mistakes
Each original cell dividing n times produces 2n events. Without back-calculation, later generations dominate the result.
Division index and proliferation index distinguish fewer responders from slower division. Reporting one alone conflates the mechanisms.
Signal approaches autofluorescence and peaks merge. Counting becomes unreliable past that point.
High CFSE concentrations are themselves antiproliferative, suppressing the response being measured. Titrate to the lowest workable level.