Data-driven discovery of chemotactic migration of bacteria via coordinate-invariant machine learning.
Chemotaxis
Machine learning
Neural letworks
Partial differential equations
Journal
BMC bioinformatics
ISSN: 1471-2105
Titre abrégé: BMC Bioinformatics
Pays: England
ID NLM: 100965194
Informations de publication
Date de publication:
24 Oct 2024
24 Oct 2024
Historique:
received:
04
04
2023
accepted:
16
09
2024
medline:
25
10
2024
pubmed:
25
10
2024
entrez:
25
10
2024
Statut:
epublish
Résumé
E. coli chemotactic motion in the presence of a chemonutrient field can be studied using wet laboratory experiments or macroscale-level partial differential equations (PDEs) (among others). Bridging experimental measurements and chemotactic Partial Differential Equations requires knowledge of the evolution of all underlying fields, initial and boundary conditions, and often necessitates strong assumptions. In this work, we propose machine learning approaches, along with ideas from the Whitney and Takens embedding theorems, to circumvent these challenges. Machine learning approaches for identifying underlying PDEs were (a) validated through the use of simulation data from established continuum models and (b) used to infer chemotactic PDEs from experimental data. Such data-driven models were surrogates either for the entire chemotactic PDE right-hand-side (black box models), or, in a more targeted fashion, just for the chemotactic term (gray box models). Furthermore, it was demonstrated that a short history of bacterial density may compensate for the missing measurements of the field of chemonutrient concentration. In fact, given reasonable conditions, such a short history of bacterial density measurements could even be used to infer chemonutrient concentration. Data-driven PDEs are an important modeling tool when studying Chemotaxis at the macroscale, as they can learn bacterial motility from various data sources, fidelities (here, computational models, experiments) or coordinate systems. The resulting data-driven PDEs can then be simulated to reproduce/predict computational or experimental bacterial density profile data independent of the coordinate system, approximate meaningful parameters or functional terms, and even possibly estimate the underlying (unmeasured) chemonutrient field evolution.
Sections du résumé
BACKGROUND
BACKGROUND
E. coli chemotactic motion in the presence of a chemonutrient field can be studied using wet laboratory experiments or macroscale-level partial differential equations (PDEs) (among others). Bridging experimental measurements and chemotactic Partial Differential Equations requires knowledge of the evolution of all underlying fields, initial and boundary conditions, and often necessitates strong assumptions. In this work, we propose machine learning approaches, along with ideas from the Whitney and Takens embedding theorems, to circumvent these challenges.
RESULTS
RESULTS
Machine learning approaches for identifying underlying PDEs were (a) validated through the use of simulation data from established continuum models and (b) used to infer chemotactic PDEs from experimental data. Such data-driven models were surrogates either for the entire chemotactic PDE right-hand-side (black box models), or, in a more targeted fashion, just for the chemotactic term (gray box models). Furthermore, it was demonstrated that a short history of bacterial density may compensate for the missing measurements of the field of chemonutrient concentration. In fact, given reasonable conditions, such a short history of bacterial density measurements could even be used to infer chemonutrient concentration.
CONCLUSION
CONCLUSIONS
Data-driven PDEs are an important modeling tool when studying Chemotaxis at the macroscale, as they can learn bacterial motility from various data sources, fidelities (here, computational models, experiments) or coordinate systems. The resulting data-driven PDEs can then be simulated to reproduce/predict computational or experimental bacterial density profile data independent of the coordinate system, approximate meaningful parameters or functional terms, and even possibly estimate the underlying (unmeasured) chemonutrient field evolution.
Identifiants
pubmed: 39448929
doi: 10.1186/s12859-024-05929-w
pii: 10.1186/s12859-024-05929-w
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
337Subventions
Organisme : National Science Foundation
ID : CBET-1941716
Organisme : Air Force Office of Scientific Research
ID : D FA9550-21-1-0317
Organisme : DOE/ Illinois Institute of Technology
ID : SA22-0052-S001
Informations de copyright
© 2024. The Author(s).
Références
Bhattacharjee T, Amchin DB, Ott JA, Kratz F, Datta SS. Chemotactic migration of bacteria in porous media. Biophys J. 2021;120(16):3483–97. https://doi.org/10.1016/j.bpj.2021.05.012 .
doi: 10.1016/j.bpj.2021.05.012
pubmed: 34022238
pmcid: 8391059
Tindall MJ, Maini PK, Porter SL, Armitage JP. Overview of mathematical approaches used to model bacterial chemotaxis ii: Bacterial populations. Bull Math Biol. 2008;70(6):1570–607. https://doi.org/10.1007/s11538-008-9322-5 .
doi: 10.1007/s11538-008-9322-5
pubmed: 18642047
Sourjik V, Berg HC. Receptor sensitivity in bacterial chemotaxis. Proceed Nat Acad Sci. 2002;99(1):123–7. https://doi.org/10.1073/pnas.011589998 .
doi: 10.1073/pnas.011589998
Tu Y. Quantitative modeling of bacterial chemotaxis: signal amplification and accurate adaptation. Annu Rev Biophys. 2013;42(1):337–59. https://doi.org/10.1146/annurev-biophys-083012-130358 . (PMID: 23451887).
doi: 10.1146/annurev-biophys-083012-130358
pubmed: 23451887
pmcid: 3737589
Colin R, Sourjik V. Emergent properties of bacterial chemotaxis pathway. Curr Opin Microbiol. 2017;39:24–33. https://doi.org/10.1016/j.mib.2017.07.004 .
doi: 10.1016/j.mib.2017.07.004
pubmed: 28822274
Setayeshgar S, Gear CW, Othmer HG, Kevrekidis IG. Application of coarse integration to bacterial chemotaxis. Multiscale Model Simul. 2005;4(1):307–27. https://doi.org/10.1137/030600874 .
doi: 10.1137/030600874
Siettos C. Coarse-grained computational stability analysis and acceleration of the collective dynamics of a monte carlo simulation of bacterial locomotion. Appl Math Comput. 2014;232:836–47. https://doi.org/10.1016/j.amc.2014.01.151 .
doi: 10.1016/j.amc.2014.01.151
Othmer HG, Hillen T. The diffusion limit of transport equations ii: Chemotaxis equations. SIAM J Appl Math. 2002;62(4):1222–50. https://doi.org/10.1137/S0036139900382772 .
doi: 10.1137/S0036139900382772
Lee S, Psarellis YM, Siettos CI, Kevrekidis IG. Learning black- and gray-box chemotactic PDEs/closures from agent based Monte Carlo simulation data. 2022;arXiv. https://doi.org/10.48550/ARXIV.2205.13545 . arXiv: https://arxiv.org/abs/2205.13545
Narla AV, Cremer J, Hwa T. A traveling-wave solution for bacterial chemotaxis with growth. Proc Natl Acad Sci. 2021;118(48):2105138118. https://doi.org/10.1073/pnas.2105138118 .
doi: 10.1073/pnas.2105138118
Keller EF, Segel LA. Model for chemotaxis. J Theor Biol. 1971;30(2):225–34. https://doi.org/10.1016/0022-5193(71)90050-6 .
doi: 10.1016/0022-5193(71)90050-6
pubmed: 4926701
Mesibov R, Ordal GW, Adler J. The range of attractant concentrations for bacterial chemotaxis and the threshold and size of response over this range: weber law and related phenomena. J Gen Physiol. 1973;62(2):203–23. https://doi.org/10.1085/jgp.62.2.203 .
doi: 10.1085/jgp.62.2.203
pubmed: 4578974
pmcid: 2226111
Tu Y, Shimizu TS, Berg HC. Modeling the chemotactic response of Escherichia coli to time-varying stimuli. Proc Natl Acad Sci. 2008;105(39):14855–60. https://doi.org/10.1073/pnas.0807569105 .
doi: 10.1073/pnas.0807569105
pubmed: 18812513
pmcid: 2551628
Fu X, Kato S, Long J, Mattingly HH, He C, Vural DC, Zucker SW, Emonet T. Spatial self-organization resolves conflicts between individuality and collective migration. Nat Commun. 2018;9(1):2177. https://doi.org/10.1038/s41467-018-04539-4 .
doi: 10.1038/s41467-018-04539-4
pubmed: 29872053
pmcid: 5988668
Cremer J, Honda T, Tang Y, Wong-Ng J, Vergassola M, Hwa T. Chemotaxis as a navigation strategy to boost range expansion. Nature. 2019;575(7784):658–63. https://doi.org/10.1038/s41586-019-1733-y .
doi: 10.1038/s41586-019-1733-y
pubmed: 31695195
pmcid: 6883170
Erban R, Othmer HG. From individual to collective behavior in bacterial chemotaxis. SIAM J Appl Math. 2004;65(2):361–91.
doi: 10.1137/S0036139903433232
Lee S, Kooshkbaghi M, Spiliotis K, Siettos CI, Kevrekidis IG. Coarse-scale PDES from fine-scale observations via machine learning. Chaos: Interdiscip J Nonlinear Sci. 2020;30(1):013141.
doi: 10.1063/1.5126869
Phan TV, Mattingly HH, Vo L, Marvin JS, Looger LL, Emonet T. Direct measurement of dynamic attractant gradients reveals breakdown of the patlak-keller-segel chemotaxis model. 2023; bioRxiv. https://doi.org/10.1101/2023.06.01.543315 . https://www.biorxiv.org/content/early/2023/06/05/2023.06.01.543315.full.pdf
Rico-Martinez R, Anderson JS, Kevrekidis IG. Continuous-time Nonlinear Signal Processing: A Neural Network Based Approach for Gray Box Identification. 1994; pp. 596–605. https://www.scopus.com/inward/record.uri?eid=2-s2.0-0028723099&partnerID=40 &md5=88de0ca8a967a54274c853e11a84d03f
González-García R, Rico-Martínez R, Kevrekidis IG. Identification of distributed parameter systems: A neural net based approach. Computers & Chemical Engineering. 1998;22, 965–68. https://doi.org/10.1016/S0098-1354(98)00191-4 .
Galaris E, Fabiani G, Gallos I, Kevrekidis I, Siettos C. Numerical bifurcation analysis of pdes from lattice boltzmann model simulations: a parsimonious machine learning approach. J Sci Comput. 2022;92(2):34. https://doi.org/10.1007/s10915-022-01883-y .
doi: 10.1007/s10915-022-01883-y
Raissi M, Perdikaris P, Karniadakis GE. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J Comput Phys. 2019;378:686–707. https://doi.org/10.1016/j.jcp.2018.10.045 .
doi: 10.1016/j.jcp.2018.10.045
Raissi M, Karniadakis GE. Hidden physics models: machine learning of nonlinear partial differential equations. J Comput Phys. 2018;357:125–41. https://doi.org/10.1016/j.jcp.2017.11.039 .
doi: 10.1016/j.jcp.2017.11.039
Yang L, Meng X, Karniadakis GE. B-pinns: Bayesian physics-informed neural networks for forward and inverse pde problems with noisy data. J Comput Phys. 2021;425:109913. https://doi.org/10.1016/j.jcp.2020.109913 .
doi: 10.1016/j.jcp.2020.109913
Chen RTQ, Rubanova Y, Bettencourt J, Duvenaud D. Neural Ordinary Differential Equations; 2019. arXiv: 1806.07366 .
LeCun Y, Bengio Y. Convolutional Networks for Images, Speech, and Time Series. 1998;pp. 255–258. MIT Press, Cambridge, MA, USA.
Rao C, Ren P, Liu Y, Sun H. Discovering Nonlinear PDEs from Scarce Data with Physics-encoded Learning. 2022. arXiv: 2201.12354 .
Brunton SL, Proctor JL, Kutz JN. Discovering governing equations from data by sparse identification of nonlinear dynamical systems. Proc Natl Acad Sci. 2016;113(15):3932–7. https://doi.org/10.1073/pnas.1517384113 .
doi: 10.1073/pnas.1517384113
pubmed: 27035946
pmcid: 4839439
Vlachas PR, Byeon W, Wan ZY, Sapsis TP, Koumoutsakos P. Data-driven forecasting of high-dimensional chaotic systems with long short-term memory networks. Proceed Royal Soc A: Math, Phys Eng Sci. 2018;474(2213):20170844. https://doi.org/10.1098/rspa.2017.0844 .
doi: 10.1098/rspa.2017.0844
Vlachas PR, Arampatzis G, Uhler C, Koumoutsakos P. Multiscale simulations of complex systems by learning their effective dynamics. Nat Mach Intell. 2022;4(4):359–66. https://doi.org/10.1038/s42256-022-00464-w .
doi: 10.1038/s42256-022-00464-w
Yu J, Lu L, Meng X, Karniadakis GE. Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems. Comput Methods Appl Mech Eng. 2022;393:114823. https://doi.org/10.1016/j.cma.2022.114823 .
doi: 10.1016/j.cma.2022.114823
Li J, Kevrekidis PG, Gear CW, Kevrekidis IG. Deciding the nature of the coarse equation through microscopic simulations: the baby-bathwater scheme. Multiscale Model Simul. 2003;1(3):391–407. https://doi.org/10.1137/S1540345902419161 .
doi: 10.1137/S1540345902419161
Whitney H. Differentiable manifolds. Ann Math. 1936;37(3):645–80.
doi: 10.2307/1968482
Takens F. Detecting strange attractors in turbulence. In: Rand D, Young L-S, editors. Dynamical Systems and Turbulence, Warwick 1980. Berlin, Heidelberg: Springer; 1981. p. 366–81.
doi: 10.1007/BFb0091924
Stark J, Broomhead DS, Davies ME, Huke J. Takens embedding theorems for forced and stochastic systems. Nonlinear Anal: Theory, Methods Appl. 1997;30(8):5303–14.
doi: 10.1016/S0362-546X(96)00149-6
Stark J. Delay embeddings for forced systems. IIS deterministic forcing. J Nonlinear Sci. 1999;9(3):255–332. https://doi.org/10.1007/s003329900072 .
doi: 10.1007/s003329900072
Stark J, Broomhead DS, Davies ME, Huke J. Delay embeddings for forced systems II. stochastic forcing. J Nonlinear Sci. 2003;13(6):519–77. https://doi.org/10.1007/s00332-003-0534-4 .
doi: 10.1007/s00332-003-0534-4
Sauer T, Yorke JA, Casdagli M. Embedology. J Stat Phys. 1991;65(3):579–616. https://doi.org/10.1007/BF01053745 .
doi: 10.1007/BF01053745
Packard NH, Crutchfield JP, Farmer JD, Shaw RS. Geometry from a time series. Phys Rev Lett. 1980;45:712–6. https://doi.org/10.1103/PhysRevLett.45.712 .
doi: 10.1103/PhysRevLett.45.712
Aeyels D. Generic observability of differentiable systems. SIAM J Control Optim. 1981;19(5):595–603. https://doi.org/10.1137/0319037 .
doi: 10.1137/0319037
Altaf MU, Titi ES, Gebrael T, Knio OM, Zhao L, McCabe MF, Hoteit I. Downscaling the 2d bénard convection equations using continuous data assimilation. Comput Geosci. 2017;21(3):393–410. https://doi.org/10.1007/s10596-017-9619-2 .
doi: 10.1007/s10596-017-9619-2
Farhat A, Lunasin E, Titi ES. Continuous data assimilation for a 2d bénard convection system through horizontal velocity measurements alone. J Nonlinear Sci. 2017;27(3):1065–87. https://doi.org/10.1007/s00332-017-9360-y .
doi: 10.1007/s00332-017-9360-y
Bar-Sinai Y, Hoyer S, Hickey J, Brenner MP. Learning data-driven discretizations for partial differential equations. Proc Natl Acad Sci. 2019;116(31):15344–9. https://doi.org/10.1073/pnas.1814058116 .
doi: 10.1073/pnas.1814058116
pubmed: 31311866
pmcid: 6681734
Wang B, Chen T. Gaussian process regression with multiple response variables. Chemom Intell Lab Syst. 2015;142:159–65. https://doi.org/10.1016/j.chemolab.2015.01.016 .
doi: 10.1016/j.chemolab.2015.01.016
Luk K, Grosse R. A Coordinate-Free Construction of Scalable Natural Gradient;2020. https://openreview.net/forum?id=H1lBYCEFDB
Weiler M, Forré P, Verlinde E, Welling M. Coordinate Independent Convolutional Networks–Isometry and Gauge Equivariant Convolutions on Riemannian Manifolds. 2021.
Flanders H. Differential Forms with Applications to the Physical Sciences. 2nd ed. New York: Dover Books on Advanced Mathematics. Dover Publications Inc; 1989.
Lee J. Manifolds and Differential Geometry. Graduate Studies in Mathematics, vol. 107. American Mathematical Society, Providence, Rhode Island. 2009. https://doi.org/10.1090/gsm/107 . http://www.ams.org/gsm/107 Accessed 2022-04-19.
Rasmussen CE, Williams CKI. Gaussian Processes for Machine Learning. Cambridge, MA: MIT Press; 2006.
Ghorbani A, Abid A, Zou J. Interpretation of neural networks is fragile. Proceed AAAI Conf Artif Intell. 2019;33(01):3681–8. https://doi.org/10.1609/aaai.v33i01.33013681 .
doi: 10.1609/aaai.v33i01.33013681
Multiphysics C. Introduction to comsol multiphysics®. COMSOL Multiphysics, Burlington, MA, accessed Feb. 1998;9, 2018.
Cybenko G. Approximation by superpositions of a sigmoidal function. Math Control Signals Syst. 1989;2(4):303–14.
doi: 10.1007/BF02551274
Kingma DP, Ba J. Adam: A Method for Stochastic Optimization. 2014; arXiv. https://doi.org/10.48550/ARXIV.1412.6980 . arXiv: https://arxiv.org/abs/1412.6980
Shampine LF, Reichelt MW. The matlab ode suite. SIAM J Sci Comput. 1997;18(1):1–22. https://doi.org/10.1137/S1064827594276424 .
doi: 10.1137/S1064827594276424
Savitzky A, Golay MJE. Smoothing and differentiation of data by simplified least squares procedures. Anal Chem. 1964;36(8):1627–39. https://doi.org/10.1021/ac60214a047 .
doi: 10.1021/ac60214a047
Getreuer P. A Survey of Gaussian Convolution Algorithms. Image Process Line. 2013;3:286–310. https://doi.org/10.5201/ipol.2013.87 .
doi: 10.5201/ipol.2013.87
Fasshauer GE. Meshfree Approximation Methods with Matlab. World Scientific, Illinois Institute of Technology, USA);2007. https://doi.org/10.1142/6437 . https://www.worldscientific.com/doi/pdf/10.1142/6437 .
Halko N, Martinsson PG, Tropp JA. Finding structure with randomness: Probabilistic algorithms for constructing approximate matrix decompositions;2010. arXiv: 0909.4061 .
Thiem TN, Kemeth FP, Bertalan T, Laing CR, Kevrekidis IG. Global and local reduced models for interacting, heterogeneous agents. Chaos: Interdiscip J Nonlinear Sci. 2021;31(7):073139. https://doi.org/10.1063/5.0055840 .
doi: 10.1063/5.0055840