A functional equation from cell kinetics (Q1198693)

The author analyzes a model of cell population growth with the following properties: (i) an individual cell grows in size as it transits the cell cycle and then divides into two daughter cells with possibly unequal sizes; (ii) the initial size of a daughter cell determines its size at mitosis through an increasing function \(\varphi\); (iii) the initial size of a daughter cell determines its cycle time through a decreasing function \(\psi\). The model takes the form of the functional equation \[ m(t,x)=2 \int^ \infty_ 0 k(x,\xi)m(t-q(x),\xi)d\xi, \] \[ \text{where } \qquad q(x)=\psi(\varphi^{-1}(x)) \quad \text{ and }\quad k(x,\xi)=[Q^{- 1}]'(x) f(Q^{-1}(x)\mid\xi), \] with \(f(y\mid x)\) the conditional probability density function for the size \(y\) of a daughter cell given that its mother cell had size \(x\). The author uses the theory of positive irreducible semigroups of linear operators to prove that the solutions have the following asymptotic behavior: there exists \(\lambda>0\) and \(h\in L^ 1(\mathbb{R}_ +)\) such that \(e^{- \lambda t} m(t,\cdot)\to Ch(\cdot)\) in \(L^ 1(\mathbb{R}_ +)\) as \(t\to\infty\), where \(C\) depends on the initial data. The model generalizes an earlier model of cell population growth of \textit{O. Arino} and \textit{M. Kimmel} [SIAM J. Appl. Math. 47, 128-145 (1987; Zbl 0632.92015)], which made more restrictive assumptions on the size range of individual cells.