Neural network approach to data-driven estimation of chemotactic sensitivity in the Keller-Segel model.

We consider the mathematical model of chemotaxis introduced by Patlak, Keller, and Segel. Aggregation and progression waves are present everywhere in the population dynamics of chemotactic cells. Aggregation originates from the chemotaxis of mobile cells, where cells are attracted to migrate to higher concentrations of the chemical signal region produced by themselves. The neural net can be used to find the approximate solution of the PDE. We proved that the error, the difference between the actual value and the predicted value, is bound to a constant multiple of the loss we are learning. Also, the Neural Net approximation can be easily applied to the inverse problem. It was confirmed that even when the coefficient of the PDE equation was unknown, prediction with high accuracy was achieved.

[1]  Juan Soler,et al.  Asymptotic Behavior of an Initial-Boundary Value Problem for the Vlasov-Poisson-Fokker-Planck System , 1997, SIAM J. Appl. Math..

[2]  P. Devreotes,et al.  Eukaryotic Chemotaxis: Distinctions between Directional Sensing and Polarization* , 2003, Journal of Biological Chemistry.

[3]  François Bouchut,et al.  On long time asymptotics of the Vlasov-Fokker-Planck equation and of the Vlasov-Poisson-Fokker-Planck system with Coulombic and Newtonian potentials , 1995, Differential and Integral Equations.

[4]  Arnulf Jentzen,et al.  Solving high-dimensional partial differential equations using deep learning , 2017, Proceedings of the National Academy of Sciences.

[5]  L. Segel,et al.  Initiation of slime mold aggregation viewed as an instability. , 1970, Journal of theoretical biology.

[6]  Xin Li,et al.  Simultaneous approximations of multivariate functions and their derivatives by neural networks with one hidden layer , 1996, Neurocomputing.

[7]  L. Segel,et al.  Model for chemotaxis. , 1971, Journal of theoretical biology.

[8]  Randall J. LeVeque,et al.  Clawpack: building an open source ecosystem for solving hyperbolic PDEs , 2016, PeerJ Comput. Sci..

[9]  Paris Perdikaris,et al.  Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations , 2019, J. Comput. Phys..

[10]  C. Patlak.  Random walk with persistence and external bias , 1953 .

[11]  L. G. Stern,et al.  Fractional step methods applied to a chemotaxis model , 2000, Journal of mathematical biology.

[12]  J. Sherratt.  Chemotaxis and chemokinesis in eukaryotic cells: the Keller-Segel equations as an approximation to a detailed model. , 1994, Bulletin of Mathematical Biology.

[13]  Hyeontae Jo,et al.  Deep Neural Network Approach to Forward-Inverse Problems , 2019, Networks Heterog. Media.

[14]  Paris Perdikaris,et al.  Physics Informed Deep Learning (Part II): Data-driven Discovery of Nonlinear Partial Differential Equations , 2017, ArXiv.

[15]  D. Zhelev,et al.  Controlled pseudopod extension of human neutrophils stimulated with different chemoattractants. , 2004, Biophysical journal.

[16]  Jae Yong Lee,et al.  Trend to equilibrium for the kinetic Fokker-Planck equation via the neural network approach , 2019, Journal of Computational Physics.

[17]  L. Segel,et al.  Traveling bands of chemotactic bacteria: a theoretical analysis. , 1971, Journal of theoretical biology.

[18]  A Goldbeter,et al.  A mechanism for exact sensory adaptation based on receptor modification. , 1986, Journal of Theoretical Biology.

[19]  J. Bourgain,et al.  Fourier transform restriction phenomena for certain lattice subsets and applications to nonlinear evolution equations , 1993 .