# American Institute of Mathematical Sciences

October  2020, 25(10): 3931-3961. doi: 10.3934/dcdsb.2020078

## Estimating the division rate from indirect measurements of single cells

 1 Sorbonne Université, Inria, Université Paris-Diderot, CNRS, Laboratoire Jacques-Louis Lions, 75005 Paris, France 2 Laboratoire de Mathematiques d'Orsay, Univ. Paris-Sud, CNRS, Université Paris-Saclay, 91405 Orsay, France 3 Laboratoire Jean Perrin, Sorbonne Université, 75005 Paris, France

* Corresponding author

Received  June 2019 Revised  December 2019 Published  January 2020

Fund Project: The first author has been supported by the ERC Starting Grant SKIPPERAD (number 306321). The second author was on leave ("délégation") at the french National Research Centre for Science (CNRS) during the finalization of this work

Is it possible to estimate the dependence of a growing and dividing population on a given trait in the case where this trait is not directly accessible by experimental measurements, but making use of measurements of another variable? This article adresses this general question for a very recent and popular model describing bacterial growth, the so-called incremental or adder model. In this model, the division rate depends on the increment of size between birth and division, whereas the most accessible trait is the size itself. We prove that estimating the division rate from size measurements is possible, we state a reconstruction formula in a deterministic and then in a statistical setting, and solve numerically the problem on simulated and experimental data. Though this represents a severely ill-posed inverse problem, our numerical results prove to be satisfactory.

Citation: Marie Doumic, Adélaïde Olivier, Lydia Robert. Estimating the division rate from indirect measurements of single cells. Discrete & Continuous Dynamical Systems - B, 2020, 25 (10) : 3931-3961. doi: 10.3934/dcdsb.2020078
##### References:

show all references

##### References:
Protocol 1 – Reconstruction of $B$ when both $U_{B,x}$ and $\mathcal{L}_B$ are (almost) exactly known. The oracle choice for $h$ gives us the value 1/4.75
Results of Protocols 1 and 2. $x$ stands for size, $\xi$ for frequency and $a$ for increment of size. Estimation of the division rate $B(a) = a^2$ in function of the increment of size $a$ (Subfigure 2g), and of all the intermediate functions necessary (Subfigures 2a to 2b)
Protocol 2 – Reconstruction of $B$ when $U_{B,x}$ is (almost) exactly known but not $\mathcal{L}_B$. The oracle choice for $h$ gives us the value 1/5
Protocol 3 – Reconstruction of $B$ when $U_{B,x}$ is reconstructed from $X_1,\ldots,X_n$ i.i.d. $\sim$ $U_{B,x}$ but $\mathcal{L}_B$ is (almost) exactly known. The oracle choice for $h_3$ gives us values that range between $1/3.25$ for $n = 500$ and $1/4.75$ for $n = 50\; 000$. We set $\varpi_n = 1/n$
Results of Protocol 3 for $n = 2000$ and $M = 100$ Monte Carlo samples. ($x$ stands for size, $\xi$ for frequency and $a$ for increment of size). Estimation of the division rate $B(a) = a^2$ in function of the increment of size $a$ (Subfigure 5e), and of intermediate functions (Subfigures 5a to 5d). In beige, the zone where 95% of the 100 Monte Carlo samples lie
Results of Protocols 3 (Left) and 4 (Right), – Estimation of the division rate $B(a) = a^2$ in function of the increment of size $a$ for different $n$ (from up to bottom, 500, 5 000, 10 000) and $M = 100$ Monte Carlo samples (the beige zone representing the zone of 95% of the 100 reconstructions). We see that the result improves for $x\leq\approx 2$ when $n$ increases, but remains very poor for larger $x$, the difference between the blue curve (Protocol 1 or 2) and the beige zone showing the influence of the sampling noise. However, Protocol 4 does not significantly worsen the results of Protocol 3
Results of Protocols 3 and 4 – Reduction of the mean error over $M = 100$ samples (in log-scale) in function of the sample size (from $n = 500$ to $n = 50\; 000$). Empirical errors are computed over the following regular grids: (a)-(e) $[0;6]$, $\Delta x = \tfrac{6}{500}$; (f)-(h) $[-10;10]$, $\Delta \xi = 0.05$; (i)-(j) $[0;2.25]$, $\Delta a = \tfrac{1}{\sqrt{n}}$; (k) $[0;2]$, $\Delta a = \tfrac{1}{\sqrt{n}}$
Protocol 4 – Reconstruction of $B$ when both $U_{B,x}$ and $\mathcal{L}_B$ are reconstructed from $X_1,\ldots,X_n$ i.i.d. $\sim$ $U_{B,x}$. The parameter $h_1$ is automatically chosen by the kernel smoothing function ${\text ksdensity}$; $h_2$ is deduced from $h_1$. The oracle choice for $h_3$ gives us values that range between $1/3.25$ for $n = 500$ and $1/4.5$ for $n = 50\; 000$. We set $\varpi_n = 1/n$
Results of Protocol 4 for $n = 2000$ and $M = 100$ Monte Carlo samples, with an inital division rate $B(a) = a^2$. From (a) to (j): successive steps of the Protocol 4. We see that the main errors come from the estimation of the Fourier transform (Subfigures 9g and 9h)
Testing the procedure on experimental data

Initial step: estimation of the size distribution.

Speed of convergence of each step of Protocol 4 in a log-log scale
Testing the procedure on experimental data

Final step: estimation of the increment-structured division rate.

Errors of Protocols 1 and 2 for the intermediate steps
 Reconstruction of $\mathcal L_B$ $\mathcal G_B$ $\mathcal G^*_B$ Numerical [0;6] [0;6] [-50;50] sampling $\Delta x = \tfrac{6}{500}$ $\Delta x = \tfrac{6}{500}$ $\Delta \xi = 0.05$ Protocol 1 - - - Protocol 2 0.0478 0.0417 0.0417
 Reconstruction of $\mathcal L_B$ $\mathcal G_B$ $\mathcal G^*_B$ Numerical [0;6] [0;6] [-50;50] sampling $\Delta x = \tfrac{6}{500}$ $\Delta x = \tfrac{6}{500}$ $\Delta \xi = 0.05$ Protocol 1 - - - Protocol 2 0.0478 0.0417 0.0417
Errors of Protocols 1 and 2 for $B$ in function of the numerical sampling
 Reconstruction of $B$ $B$ Numerical [0;2] [0;2.5 ] sampling $\Delta a = 0.01$ $\Delta a = 0.01$ Protocol 1 0.0730 0.2065 Protocol 2 0.0849 0.1321
 Reconstruction of $B$ $B$ Numerical [0;2] [0;2.5 ] sampling $\Delta a = 0.01$ $\Delta a = 0.01$ Protocol 1 0.0730 0.2065 Protocol 2 0.0849 0.1321
 [1] Gunther Uhlmann, Jian Zhai. Inverse problems for nonlinear hyperbolic equations. Discrete & Continuous Dynamical Systems - A, 2021, 41 (1) : 455-469. doi: 10.3934/dcds.2020380 [2] H. M. Srivastava, H. I. Abdel-Gawad, Khaled Mohammed Saad. Oscillatory states and patterns formation in a two-cell cubic autocatalytic reaction-diffusion model subjected to the Dirichlet conditions. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020433 [3] Yi-Hsuan Lin, Gen Nakamura, Roland Potthast, Haibing Wang. Duality between range and no-response tests and its application for inverse problems. Inverse Problems & Imaging, , () : -. doi: 10.3934/ipi.2020072 [4] Kha Van Huynh, Barbara Kaltenbacher. Some application examples of minimization based formulations of inverse problems and their regularization. Inverse Problems & Imaging, , () : -. doi: 10.3934/ipi.2020074 [5] Aihua Fan, Jörg Schmeling, Weixiao Shen. $L^\infty$-estimation of generalized Thue-Morse trigonometric polynomials and ergodic maximization. Discrete & Continuous Dynamical Systems - A, 2021, 41 (1) : 297-327. doi: 10.3934/dcds.2020363 [6] Wenbin Li, Jianliang Qian. Simultaneously recovering both domain and varying density in inverse gravimetry by efficient level-set methods. Inverse Problems & Imaging, , () : -. doi: 10.3934/ipi.2020073 [7] Monia Capanna, Jean C. Nakasato, Marcone C. Pereira, Julio D. Rossi. Homogenization for nonlocal problems with smooth kernels. Discrete & Continuous Dynamical Systems - A, 2020  doi: 10.3934/dcds.2020385 [8] Laurence Cherfils, Stefania Gatti, Alain Miranville, Rémy Guillevin. Analysis of a model for tumor growth and lactate exchanges in a glioma. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020457 [9] Vieri Benci, Sunra Mosconi, Marco Squassina. Preface: Applications of mathematical analysis to problems in theoretical physics. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020446 [10] Weiwei Liu, Jinliang Wang, Yuming Chen. Threshold dynamics of a delayed nonlocal reaction-diffusion cholera model. Discrete & Continuous Dynamical Systems - B, 2020  doi: 10.3934/dcdsb.2020316 [11] Siyang Cai, Yongmei Cai, Xuerong Mao. A stochastic differential equation SIS epidemic model with regime switching. Discrete & Continuous Dynamical Systems - B, 2020  doi: 10.3934/dcdsb.2020317 [12] Lars Grüne, Matthias A. Müller, Christopher M. Kellett, Steven R. Weller. Strict dissipativity for discrete time discounted optimal control problems. Mathematical Control & Related Fields, 2020  doi: 10.3934/mcrf.2020046 [13] Yining Cao, Chuck Jia, Roger Temam, Joseph Tribbia. Mathematical analysis of a cloud resolving model including the ice microphysics. Discrete & Continuous Dynamical Systems - A, 2021, 41 (1) : 131-167. doi: 10.3934/dcds.2020219 [14] Shiqiu Fu, Kanishka Perera. On a class of semipositone problems with singular Trudinger-Moser nonlinearities. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020452 [15] Zhouchao Wei, Wei Zhang, Irene Moroz, Nikolay V. Kuznetsov. Codimension one and two bifurcations in Cattaneo-Christov heat flux model. Discrete & Continuous Dynamical Systems - B, 2020  doi: 10.3934/dcdsb.2020344 [16] Shuang Chen, Jinqiao Duan, Ji Li. Effective reduction of a three-dimensional circadian oscillator model. Discrete & Continuous Dynamical Systems - B, 2020  doi: 10.3934/dcdsb.2020349 [17] Zhiyan Ding, Qin Li, Jianfeng Lu. Ensemble Kalman Inversion for nonlinear problems: Weights, consistency, and variance bounds. Foundations of Data Science, 2020  doi: 10.3934/fods.2020018 [18] Barbora Benešová, Miroslav Frost, Lukáš Kadeřávek, Tomáš Roubíček, Petr Sedlák. An experimentally-fitted thermodynamical constitutive model for polycrystalline shape memory alloys. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020459 [19] Cuicui Li, Lin Zhou, Zhidong Teng, Buyu Wen. The threshold dynamics of a discrete-time echinococcosis transmission model. Discrete & Continuous Dynamical Systems - B, 2020  doi: 10.3934/dcdsb.2020339 [20] Giuseppina Guatteri, Federica Masiero. Stochastic maximum principle for problems with delay with dependence on the past through general measures. Mathematical Control & Related Fields, 2020  doi: 10.3934/mcrf.2020048

2019 Impact Factor: 1.27