pith. sign in

arxiv: 2404.19367 · v2 · submitted 2024-04-30 · 🧮 math.ST · stat.TH

Parametric estimation and LAN property of the birth-death-move process with mutations

Lisa Balsollier (LMJL , SAIRPICO) , Fr\'ed\'eric Lavancier (CREST) This is my paper

Pith reviewed 2026-05-24 01:57 UTC · model grok-4.3

classification 🧮 math.ST stat.TH
keywords birth-death-move processlocal asymptotic normalitymaximum likelihood estimationparametric estimationparticle systemscolocalizationprotein dynamics
0
0 comments X

The pith

A parametric birth-death-move process with mutations satisfies local asymptotic normality, yielding an efficient maximum likelihood estimator with explicit asymptotic covariance.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that a Markov model of interacting marked particles undergoing births, deaths, movements, and mutations can be parametrized so that its likelihood satisfies local asymptotic normality. From this, the maximum likelihood estimator is shown to be asymptotically efficient and normally distributed, with an explicit formula for the covariance matrix. This framework is applied to track two types of proteins in living cells to measure their colocalization during exocytosis. A reader cares because it supplies the statistical tools needed to draw reliable conclusions from observations of complex particle systems in biology.

Core claim

Assuming a parametric form, the likelihood of the birth-death-move process with mutations is derived and proved to satisfy the local asymptotic normality property. Consequently, the maximum likelihood estimator is asymptotically efficient and normal with explicit covariance matrix. The required technical assumptions hold for several natural parametric specifications used in the biological application.

What carries the argument

The local asymptotic normality (LAN) property of the parametric likelihood, which establishes the asymptotic efficiency of the MLE.

If this is right

  • The MLE achieves the Cramér-Rao lower bound asymptotically.
  • An explicit expression for the asymptotic covariance matrix is available for inference.
  • The LAN property holds for natural parametric families applied to protein data.
  • The model enables quantification of colocalization in the exocytosis process.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The LAN result may support likelihood ratio tests for specific mutation or interaction parameters.
  • Similar LAN derivations could apply to related spatial birth-death processes without mutations.
  • The explicit covariance could guide experimental design for observing particle systems at sufficient resolution.

Load-bearing premise

The chosen parametric specifications satisfy the differentiability and integrability conditions needed to establish the LAN property.

What would settle it

If for a natural parametric family the log-likelihood does not admit the LAN expansion around the true parameter, then the efficiency claim would fail.

Figures

Figures reproduced from arXiv: 2404.19367 by Fr\'ed\'eric Lavancier (CREST), Lisa Balsollier (LMJL, SAIRPICO).

Figure 1
Figure 1. Figure 1: Example of the density (20) for the birth of a new Langerin particle, given the Rab-11 particles already present (shown as red dots). pre-jump configuration x ∈ E, they are generated according to the density k L β (x, .) defined for all z ∈ Λ by k L β (x, z) = p nR(x) nXR(x) i=1 1 2πσ2 exp  −∥z − z R i ∥ 2 2σ 2  + (1 − p) |Λ| 1Λ(z), (20) where nR(x) denotes the number of existing Rab-11 particles and z R… view at source ↗
Figure 2
Figure 2. Figure 2: Estimation results for (p, log σ) = (0.2, 1.34) by MLE, based on 500 simulations of the BDMM process specified in Section 4.1. In red: 95% Gaussian confidence ellipsoid, based on the bivariate Gaussian distribution centred at the mean of all estimations, and with covariance matrix the mean of all estimates of the asymptotic covariance matrix J −1 (ϑ). 1.34, on a time interval similar to the data studied in… view at source ↗
Figure 3
Figure 3. Figure 3: Left: a frame from the raw video sequence analysed in Section [PITH_FULL_IMAGE:figures/full_fig_p024_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Likelihood of the BDMM model for the considered dataset, as [PITH_FULL_IMAGE:figures/full_fig_p025_4.png] view at source ↗
read the original abstract

A birth-death-move process with mutations is a Markov model for a system of marked particles in interaction, that move over time, with births and deaths. In addition the mark of each particle may also change, which constitutes a mutation. Assuming a parametric form for this model, we derive its likelihood expression and prove its local asymptotic normality. The efficiency and asymptotic distribution of the maximum likelihood estimator, with an explicit expression of its covariance matrix, is deduced. The underlying technical assumptions are showed to be satisfied by several natural parametric specifications. As an application, we leverage this model to analyse the joint dynamics of two types of proteins in a living cell, that are involved in the exocytosis process. Our approach enables to quantify the so-called colocalization phenomenon, answering an important question in microbiology.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 2 minor

Summary. The manuscript develops a parametric version of the birth-death-move process with mutations for marked interacting particles. It derives the likelihood, proves the local asymptotic normality (LAN) property under stated regularity conditions, deduces the asymptotic efficiency and normality of the MLE together with an explicit asymptotic covariance matrix, verifies that the technical assumptions hold for several natural parametric families, and applies the model to quantify protein colocalization in a cellular exocytosis process.

Significance. If the LAN property and MLE results hold, the work supplies a self-contained asymptotic theory for inference on this class of particle systems, including an explicit covariance that facilitates practical use. The explicit verification of assumptions for biologically motivated parametric specifications and the concrete application to colocalization data add direct value for statistical analysis in microbiology.

minor comments (2)
  1. [Abstract] Abstract: the sentence 'The efficiency and asymptotic distribution of the maximum likelihood estimator, with an explicit expression of its covariance matrix, is deduced' has a subject-verb agreement issue that reduces readability.
  2. [Model definition] The definition of the parametric rates (birth, death, move, mutation) would benefit from an explicit display of the parameter vector and its dimension early in the model section to aid readers.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive assessment of the manuscript, including the summary of its contributions to the LAN property, MLE efficiency, verification of assumptions, and the biological application. We appreciate the recommendation for minor revision. No specific major comments were provided in the report.

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper assumes a parametric form for the birth-death-move process, explicitly derives the likelihood expression, and establishes the LAN property via standard contiguity and quadratic expansion arguments under stated regularity conditions. It then verifies those conditions for the chosen parametric families and deduces MLE properties from LAN. No step reduces a claimed result to a fitted input by construction, no self-citation is load-bearing for the central claims, and the technical assumptions are checked independently for the model specifications. The derivation chain is self-contained against external statistical benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

Based solely on the abstract, the central claim rests on an assumed parametric form for event rates and on technical conditions required for LAN that the authors state are met by natural choices; no explicit free parameters or invented entities are named.

free parameters (1)
  • parametric rates for birth, death, move and mutation
    The model is defined under a parametric specification whose concrete functional form is chosen by the user and must satisfy the LAN conditions.
axioms (1)
  • domain assumption Technical assumptions required for LAN are satisfied by several natural parametric specifications
    Invoked when claiming that the LAN property and explicit MLE covariance hold for the families used in the protein application.

pith-pipeline@v0.9.0 · 5669 in / 1265 out tokens · 25495 ms · 2026-05-24T01:57:20.875245+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages

  1. [1]

    B., and Ney, P

    Athreya, K. B., and Ney, P. E. Branching Processes, vol. 196. Springer Science & Business Media, 2012

    work page 2012

  2. [2]

    A generative model to simulate spatiotemporal dynamics of biomolecules in cells

    Balsollier, L., Lavancier, F., Salamero, J., and Kervrann, C. A generative model to simulate spatiotemporal dynamics of biomolecules in cells. Biological Imaging 3 (2023), e22

    work page 2023

  3. [3]

    Stochastic models for structured populations, vol

    Bansaye, V., and M ´el´eard, S. Stochastic models for structured populations, vol. 16. Springer, 2015

    work page 2015

  4. [4]

    Fast high-resolution 3D total inter- nal reflection fluorescence microscopy by incidence angle scanning and azimuthal averaging

    Boulanger, J., Gueudry, C., Munch, D., Cinquin, B., Paul- Gilloteaux, P., Bardin, S., Gu ´erin, C., Senger, F., Blan- choin, L., and Salamero, J. Fast high-resolution 3D total inter- nal reflection fluorescence microscopy by incidence angle scanning and azimuthal averaging. Proc Natl Acad Sci USA 111 , 48 (2014), 17164– 17169. 46

    work page 2014

  5. [5]

    Statistical analysis of particle trajectories in living cells

    Briane, V., Kervrann, C., and Vimond, M. Statistical analysis of particle trajectories in living cells. Phys Rev E. 97 , (6-1):062121 (2018)

    work page 2018

  6. [6]

    An overview of dif- fusion models for intracellular dynamics analysis

    Briane, V., Vimond, M., and Kervrann, C. An overview of dif- fusion models for intracellular dynamics analysis. Briefings in bioinfor- matics 21, 4 (2020), 1136–1150

    work page 2020

  7. [7]

    Strong feller properties for dis- torted brownian motion with reflecting boundary condition and an ap- plication to continuous n-particle systems with singular interactions

    Fattler, T., and Grothaus, M. Strong feller properties for dis- torted brownian motion with reflecting boundary condition and an ap- plication to continuous n-particle systems with singular interactions. Journal of Functional Analysis 246 , 2 (2007), 217–241

    work page 2007

  8. [8]

    Limit theorems for null recurrent markov processes

    Hopfner, R., and L¨ocherbach, E. Limit theorems for null recurrent markov processes. Univ., Fachbereich Mathematik, 2000

    work page 2000

  9. [9]

    Ibragimov, H. M. Statistical estimation: asymptotic theory , vol. 16. Springer Science & Business Media, 2013

    work page 2013

  10. [10]

    L., and Danuser, G

    Jaqaman, K., Loerke, D., Mettlen, M., Kuwata, H., Grin- stein, S., Schmid, S. L., and Danuser, G. Robust single-particle tracking in live-cell time-lapse sequences. Nature methods 5 , 8 (2008), 695

    work page 2008

  11. [11]

    The classification of birth and death processes

    Karlin, S., and McGregor, J. The classification of birth and death processes. Transactions of the American Mathematical Society 86 , 2 (1957), 366–400

    work page 1957

  12. [12]

    Klebaner, F. C. Introduction to stochastic calculus with applications. World Scientific Publishing Company, 2012

    work page 2012

  13. [13]

    Kutoyants, Y. A. Statistical inference for ergodic diffusion processes. Springer Science & Business Media, 2004

    work page 2004

  14. [14]

    Feller and ergodic properties of jump-move processes with applications to inter- acting particle systems

    Lavancier, F., Le Gu ´evel, R., , and Manent, E. Feller and ergodic properties of jump-move processes with applications to inter- acting particle systems. to appear in Journal of Applied Probability (arXiv:2204.02851) (2024)

    work page arXiv 2024

  15. [15]

    Lavancier, F., and Le Gu´evel, R. Spatial birth–death–move pro- cesses: Basic properties and estimation of their intensity functions.Jour- nal of the Royal Statistical Society: Series B (Statistical Methodology) 83, 4 (2021), 798–825. 47

    work page 2021

  16. [16]

    S., and Shiryaev, A

    Liptser, R. S., and Shiryaev, A. N. Statistics of random processes I: General Theory. Springer Science & Business Media, 2001

    work page 2001

  17. [17]

    LAN and LAMN for systems of interacting diffusions with branching and immigration

    Locherbach, E. LAN and LAMN for systems of interacting diffusions with branching and immigration. InAnnales de l’Institut Henri Poincare (B) Probability and Statistics (2002), vol. 38, Elsevier, pp. 59–90

    work page 2002

  18. [18]

    Likelihood ratio processes for markovian particle sys- tems with killing and jumps

    Locherbach, E. Likelihood ratio processes for markovian particle sys- tems with killing and jumps. Statistical inference for stochastic processes 5, 2 (2002), 153–177

    work page 2002

  19. [19]

    Local asymptotic mixed normality for semimartingale experiments

    Luschgy, H. Local asymptotic mixed normality for semimartingale experiments. Probability theory and related fields 92, 2 (1992), 151–176

    work page 1992

  20. [20]

    Propri´ et´ es de Feller, d’ergodicit´ e et inf´ erence non- param´ etrique du processus jump-move

    Manent, E. Propri´ et´ es de Feller, d’ergodicit´ e et inf´ erence non- param´ etrique du processus jump-move. PhD thesis, Universit´ e Rennes 2, 2023

    work page 2023

  21. [21]

    Temporal network epidemiology

    Masuda, N., and Holme, P. Temporal network epidemiology . Springer, 2017

    work page 2017

  22. [22]

    Statistical analysis of a spatial birth- and-death process model with a view to modelling linear dune fields

    Møller, J., and Sørensen, M. Statistical analysis of a spatial birth- and-death process model with a view to modelling linear dune fields. Scandinavian journal of statistics 21 , 1 (1994), 1–19

    work page 1994

  23. [23]

    Background flu- orescence estimation and vesicle segmentation in live cell imaging with conditional random fields

    P´ecot, T., Bouthemy, P., Boulanger, J., Chessel, A., Bardin, S., Salamero, J., and Kervrann, C. Background flu- orescence estimation and vesicle segmentation in live cell imaging with conditional random fields. IEEE Transactions on Image Processing 24 , 2 (2014), 667–680

    work page 2014

  24. [24]

    Individual-Based Methods in Forest Ecology and Management

    Pommerening, A., and Grabarnik, P. Individual-Based Methods in Forest Ecology and Management . Springer, 2019

    work page 2019

  25. [25]

    Spatial birth and death processes

    Preston, C. Spatial birth and death processes. Advances in applied probability 7, 3 (1975), 371–391

    work page 1975

  26. [26]

    Gibbs point processes for studying the development of spatial-temporal stochastic processes

    Renshaw, E., and S¨arkk¨a, A. Gibbs point processes for studying the development of spatial-temporal stochastic processes. Computational statistics & data analysis 36 , 1 (2001), 85–105. 48

    work page 2001

  27. [27]

    Analysis of the appearance and disappearance of point objects over time

    Sadahiro, Y. Analysis of the appearance and disappearance of point objects over time. International Journal of Geographical Information Science 33, 2 (2019), 215–239

    work page 2019

  28. [28]

    A new metric between distributions of point processes

    Schuhmacher, D., and Xia, A. A new metric between distributions of point processes. Advances in applied probability 40, 3 (2008), 651–672. 49

    work page 2008