Limit theorems for bifurcating Markov chains. Application to the detection of cellular aging (Q2475031)

The author proposes a general method to study dependent data in a binary tree where an individual in one generation gives rise to two different offspring, one of type 0 and another of type 1, in the next generation. For any specific characteristic of these offspring, it is assumed that the characteristic depends on its ancestors only through the mother's characteristic. Let \(X_i\) denote the growth rate of individual \(i\) and \(``n"\) denote the mother of the offspring \(``2n"\) and \(``2n+1".\) Suppose \(X_1,\) the growth rate of the ancestor, has distribution \(\nu\) and for all \(n \geq 1, X_{2n}= \alpha_0 X_n + \beta_0 +\epsilon_{2n}\) and \(X_{2n+1}= \alpha_1 X_n + \beta_1 + \epsilon_{2n+1}\), where \(\alpha_0, \alpha_1 \in (-1,1),\;\beta_0,\beta_1 \in R\), and \(((\epsilon_{2n}, \epsilon_{2n+1}), n \geq 1)\) forms a sequence of bivariate Gaussian random vectors with \(E(\epsilon_{2n}= E(\epsilon_{2n+1}) = 0, Var (\epsilon _{2n})= Var(\epsilon_{2n+1})= \sigma^2 > 0,\) and \(Cov(\epsilon_{2n}, \epsilon_{2n+1})= \rho \sigma^2\) with \(\rho \in (-1,1).\) Such a process is called bifurcating Markov chain. The author studies law of large numbers and central limit theorems for such processes and investigates the problems of estimating the 4-dimensional parameter \(\theta =(\alpha_0, \beta_0, \alpha_1, \alpha_1, \beta_1), \rho\) and \(\sigma^2,\) and testing the hypthesis \(H_0: (\alpha_0, \beta_0) = \alpha_1, \beta_1)\) against the alternative \(H_1:(\alpha_0, \beta_0) \neq (\alpha_1, \beta_1).\) An application to detect celluar aging in Escherichia Coli is discussed.