Division-Linked Expansion and Daughter-Cell Accounting

Why a Division Event Is Not the Same as Net Growth

A dividing cell disappears as a single cell and is replaced by its daughters. That event can expand a population, maintain the number of cells in a compartment, or reduce that compartment, depending on the daughters’ fates. Consequently, a rate describing how often cells divide is not automatically the rate at which cell number increases.

This distinction matters when constructing cell-population models from biological descriptions. “Cells divide at a rate of 0.2 per day” specifies an event frequency per cell. To determine the change in a particular population, we must also ask how many daughters enter that population and whether the dividing parent is counted there before the event.

The rate \(\beta\) alone does not specify the net contribution. For example, if each event replaces one parent with two viable daughters in the same population, the population gains one cell per event. If each event replaces a parent with one daughter in that population, its count is unchanged by division, even though divisions are occurring.

From Daughter Fates to a Count Equation

Begin with one compartment containing \(C\) cells. Suppose each cell divides with per-cell event rate \(\beta\), and let \(K\) be the number of daughters from an event that are counted in this same compartment. The parent is in the compartment before division and is removed when it divides. Thus, the change in the compartment’s count at one event is \(K-1\), not \(K\).

If daughter number is random, use its expected value, \(E[K]\), to obtain the expected count change per event. The resulting net division contribution is the event rate multiplied by the expected count change per event.

This equation assumes that the cells in the compartment share the same division-event rate, each division removes one counted parent, and the daughter fates are described by the specified distribution of \(K\). It gives the expected rate of change; an individual cell’s event is discrete, while the equation describes population-level mean change. Other processes, such as cell loss or transitions not included in the daughter fates, must be accounted for separately.

Accounting When Daughters Enter Different Compartments

For multiple cell types, track daughter outcomes by destination. Let \(C_i\) be the count of the dividing-parent compartment \(i\), and let \(\beta_i\) be its division-event rate per cell. Define \(a_{ji}\) as the expected number of daughters entering compartment \(j\) per division event of a cell in compartment \(i\). A daughter that does not enter any modelled compartment contributes zero to the relevant \(a_{ji}\).

Each event removes one parent from compartment \(i\). It adds the expected number \(a_{ji}\) to compartment \(j\). The contribution of divisions of type \(i\) to compartment \(j\) is therefore:

where \(\mathbf{1}_{\{j=i\}}\) equals 1 when \(j=i\) and 0 otherwise. Summing over all dividing compartments gives the total division contribution to \(C_j\). The subtraction applies only to the parent’s compartment; it is not a loss from every destination compartment.

Derivation and a Stem–Progenitor Example

The count equation follows by considering a short interval of duration \(\Delta t\). For one cell with event rate \(\beta\), the probability of a division during that interval is \(\beta\Delta t+o(\Delta t)\), where \(o(\Delta t)\) denotes terms that vanish faster than \(\Delta t\). The expected change in the compartment count from that cell is the event probability multiplied by the expected change \(E[K]-1\). Contributions from two or more events involving that cell are of smaller order over this interval.

For \(C\) cells with the same rate and daughter-fate distribution, add the expected contributions across cells, divide by \(\Delta t\), and take the limit as \(\Delta t\) tends to zero. The terms \(o(\Delta t)/\Delta t\) vanish, giving:

This is a mean-rate derivation. It does not say that fractional numbers of cells occur in individual events; fractions arise when an expected rate averages over many possible events.

Choosing the Right Population-Growth Term

For a single compartment in which each division produces a fixed number \(m\) of counted daughters, the equation reduces to \(dC/dt=\beta(m-1)C\), if division is the only process changing the count. The division-event rate is \(\beta C\), gross daughter production is \(m\beta C\), and net production is \((m-1)\beta C\). These quantities have different interpretations even though they share the same units of cells per time, except that the event rate is measured in events per time.

A common alternative is to write an effective net-growth equation directly, such as \(dC/dt=rC\). This is a compact description of observed net change, but it does not by itself identify how many divisions occurred or what happened to the daughters. In the special case of two viable daughters always entering the same compartment, \(r=\beta\). In other cases, \(r\) cannot be read directly as the division-event rate. Explicit daughter accounting is preferable when lineage output or division fates matter.

A more detailed model can represent cell-cycle phases as separate compartments and account for cells moving through them before division. That structure can represent delays in cell-cycle progression. The daughter-accounting equations remain relevant at the division event, but a single per-cell event rate may not fully represent the timing of divisions when phase occupancy and durations are modelled explicitly.

Assumptions, Diagnostics, and Common Mistakes

The daughter-accounting terms are appropriate when the model’s compartments and event outcomes are clearly defined. In practice, check the following before interpreting a fitted or simulated growth term:

• State what is counted. Specify which viable cell types belong to each compartment. A daughter’s destination must be assigned consistently.
• Check parent removal. A dividing parent is counted before the event and must be removed from its original compartment. Do not subtract it again if the model’s event definition already handles that removal elsewhere.
• Check daughter yields. For each event type, list the possible outcomes and calculate expected daughter numbers from their probabilities. Probabilities must sum to one, and yields should agree with the outcomes described.
• Keep losses distinct. A daughter that dies immediately can be represented as not entering a viable-cell compartment. Do not also apply a separate loss to that same daughter unless the model represents a distinct later loss process.
• Check units and signs. The rate \(\beta_i\) has units of inverse time; multiplying it by a cell count and a daughter yield gives cells per time. A yield below one in a compartment can produce a negative contribution there.

One frequent modelling error is to equate gross daughter production with net population growth. If 40 cells divide per day and each produces two daughters in the same compartment, gross production is 80 daughters per day, but 40 parents are removed. The net contribution is therefore \(80-40=40\) cells per day, not 80. Conversely, if one stem daughter and one differentiated daughter are produced, the stem compartment receives one stem cell but also loses its stem parent; its division contribution is zero, while the differentiated compartment gains one cell.

Daughter accounting also has limits. It assumes that the event rate and daughter-fate distribution adequately describe division for the population being modelled. If cells within a compartment have materially different division behaviour, a single rate and average yield may conceal important structure. If fates depend on time or cellular conditions, the assumed probabilities may not remain appropriate. The resulting equations describe count changes, not the biological mechanism that determines the fates.