A composite likelihood approach to the analysis of longitudinal clonal data on multitype cellular systems under an age-dependent branching process

Department of Biostatistics and Computational Biology, University of Rochester Medical Center, 601 Elmwood Avenue, Rochester, NY 14642, USA.
Biostatistics (Impact Factor: 2.65). 01/2011; 12(1):173-91. DOI: 10.1093/biostatistics/kxq050
Source: PubMed

ABSTRACT A recurrent statistical problem in cell biology is to draw inference about cell kinetics from observations collected at discrete time points. We investigate this problem when multiple cell clones are observed longitudinally over time. The theory of age-dependent branching processes provides an appealing framework for the quantitative analysis of such data. Likelihood inference being difficult in this context, we propose an alternative composite likelihood approach, where the estimation function is defined from the marginal or conditional distributions of the number of cells of each observable cell type. These distributions have generally no closed-form expressions but they can be approximated using simulations. We construct a bias-corrected version of the estimating function, which also offers computational advantages. Two algorithms are discussed to compute parameter estimates. Large sample properties of the estimator are presented. The performance of the proposed method in finite samples is investigated in simulation studies. An application to the analysis of the generation of oligodendrocytes from oligodendrocyte type-2 astrocyte progenitor cells cultured in vitro reveals the effect of neurothrophin-3 on these cells. Our work demonstrates also that the proposed approach outperforms the existing ones.

Download full-text


Available from: Mark Noble, Feb 19, 2015
9 Reads
  • Source
    • "Chen et al (2010) investigated composite likelihood estimators. Hyrien and Zand (2008) and Hyrien, Chen and Zand (2010) proposed a method combining a mixture model and composite likelihood for CFSE-labeling data. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This article deals with quasi- and pseudo-likelihood estimation in a class of continuous-time multi-type Markov branching processes observed at discrete points in time. "Conventional" and conditional estimation are discussed for both approaches. We compare their properties and identify situations where they lead to asymptotically equivalent estimators. Both approaches possess robustness properties, and coincide with maximum likelihood estimation in some cases. Quasi-likelihood functions involving only linear combinations of the data may be unable to estimate all model parameters. Remedial measures exist, including the resort either to non-linear functions of the data or to conditioning the moments on appropriate sigma-algebras. The method of pseudo-likelihood may also resolve this issue. We investigate the properties of these approaches in three examples: the pure birth process, the linear birth-and-death process, and a two-type process that generalizes the previous two examples. Simulations studies are conducted to evaluate performance in finite samples.
    Journal of Statistical Planning and Inference 07/2011; 141(7):2209-2227. DOI:10.1016/j.jspi.2011.01.016 · 0.68 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Over the past decade, flow cytometric CFSE-labeling experiments have gained considerable popularity among experimentalists, especially immunologists and hematologists, for studying the processes of cell proliferation and cell death. Several mathematical models have been presented in the literature to describe cell kinetics during these experiments. We propose a multi-type age-dependent branching process to model the temporal development of populations of cells subject to division and death during CFSE-labeling experiments. We discuss practical implementation of the proposed model; we investigate a competing risk version of the process; and we identify the classes of cellular dependencies that may influence the expectation of the process and those that do not. An application is presented where we study the proliferation of human CD8+ T lymphocytes using our model and a competing risk branching process. The proposed model offers a widely applicable approach to the analysis of CFSE-labeling experiments. The model fitted very well our experimental data. It provided reasonable estimates of cell kinetics parameters as well as meaningful insights into the processes of cell division and cell death. In contrast, the competing risk branching process could not describe the kinetics of CD8+ T cells. This suggested that the decision of cell division or cell death may be made early in the cell cycle if not in preceding generations. Also, we show that analyses based on the proposed model are robust with respect to cross-sectional dependencies and to dependencies between fates of linearly filiated cells. This article was reviewed by Marek Kimmel, Wai-Yuan Tan and Peter Olofsson.
    Biology Direct 06/2010; 5(1):41. DOI:10.1186/1745-6150-5-41 · 4.66 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Two-type reducible age-dependent branching processes with inhomogeneous immigration are considered to describe the kinetics of renewing cell populations. This class of processes can be used to model the generation of oligodendrocytes in the central nervous system in vivo or the kinetics of leukemia cells. The asymptotic behavior of the first and second moments, including the correlation, of the process is investigated.
    01/2011; 20:81-108.
Show more