Data-driven identification of biological systems using multi-scale analysis.
Uložené v:
| Názov: | Data-driven identification of biological systems using multi-scale analysis. |
|---|---|
| Autori: | Muhammed I; Department of Mathematics, Khalifa University of Science and Technology, Abu Dhabi, United Arab Emirates., Manias DM; Department of Mathematics, Khalifa University of Science and Technology, Abu Dhabi, United Arab Emirates., Goussis DA; Department of Mechanical Engineering, Khalifa University of Science and Technology, Abu Dhabi, United Arab Emirates., Hatzikirou H; Department of Mathematics, Khalifa University of Science and Technology, Abu Dhabi, United Arab Emirates.; Biotechnology Center, Khalifa University of Science and Technology, Abu Dhabi, United Arab Emirates.; Center for Information Services and High Performance Computing, Technische Universitat Dresden, Dresden, Germany. |
| Zdroj: | PLoS computational biology [PLoS Comput Biol] 2025 Nov 06; Vol. 21 (11), pp. e1013193. Date of Electronic Publication: 2025 Nov 06 (Print Publication: 2025). |
| Spôsob vydávania: | Journal Article |
| Jazyk: | English |
| Informácie o časopise: | Publisher: Public Library of Science Country of Publication: United States NLM ID: 101238922 Publication Model: eCollection Cited Medium: Internet ISSN: 1553-7358 (Electronic) Linking ISSN: 1553734X NLM ISO Abbreviation: PLoS Comput Biol Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: San Francisco, CA : Public Library of Science, [2005]- |
| Výrazy zo slovníka MeSH: | Models, Biological* , Systems Biology*/methods, Algorithms ; Neural Networks, Computer ; Nonlinear Dynamics ; Computational Biology/methods ; Computer Simulation |
| Abstrakt: | Competing Interests: The authors have declared that no competing interests exist. Biological systems inherently exhibit multi-scale dynamics, making accurate system identification particularly challenging due to the complexity of capturing a wide time scale spectrum. Traditional methods capable of addressing this issue rely on explicit equations, limiting their applicability in cases where only observational data are available. To overcome this limitation, we propose a data-driven framework that integrates the Sparse Identification of Nonlinear Dynamics (SINDy) method, the multi scale analysis algorithm Computational Singular Perturbation (CSP) and neural networks (NNs). This framework allows the partition of the available dataset in subsets characterized by similar dynamics, so that system identification can proceed within these subsets without facing a wide time scale spectrum. Accordingly, when the full dataset does not allow SINDy to identify the proper model, CSP is employed for the generation of subsets of similar dynamics, which are then fed into SINDy. CSP requires the availability of the gradient of the vector field, which is estimated by the NNs. The framework is tested on the Michaelis-Menten model, for which various reduced models in analytic form exist at different parts of the phase space. It is demonstrated that the CSP-based data subsets allow SINDy to identify the proper reduced model in cases where the full dataset does not. In addition, it is demonstrated that the framework succeeds even in the cases where the available data set originates from stochastic versions of the Michaelis-Menten model. This framework is algorithmic, so system identification is not hindered by the dimensions of the dataset. (Copyright: © 2025 Muhammed et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.) |
| References: | Phys Life Rev. 2022 Mar;40:1-2. (PMID: 35085921) Math Biosci. 2019 Sep;315:108220. (PMID: 31255632) J R Soc Interface. 2021 Dec;18(185):20210686. (PMID: 34932929) Proc Math Phys Eng Sci. 2017 Jan;473(2197):20160446. (PMID: 28265183) FEBS J. 2014 Jan;281(2):435-63. (PMID: 24180270) Sci Rep. 2020 Nov 16;10(1):19858. (PMID: 33199734) PLoS Comput Biol. 2024 Jun 6;20(6):e1012205. (PMID: 38843305) J Theor Biol. 2022 Feb 7;534:110975. (PMID: 34883121) Mol Biol Cell. 2012 Feb;23(4):517-9. (PMID: 22337858) PLoS One. 2016 Feb 26;11(2):e0150171. (PMID: 26919740) Proc Natl Acad Sci U S A. 2016 Apr 12;113(15):3932-7. (PMID: 27035946) J Comput Phys. 2021 Oct 15;443:. (PMID: 34744183) Front Immunol. 2024 Mar 12;15:1363144. (PMID: 38533513) NPJ Syst Biol Appl. 2025 Jan 8;11(1):3. (PMID: 39779740) iScience. 2024 Feb 23;27(4):109316. (PMID: 38523784) PLoS Comput Biol. 2024 May 3;20(5):e1012088. (PMID: 38701089) Prog Biophys Mol Biol. 2008 Jan-Apr;96(1-3):60-89. (PMID: 17888502) Front Comput Neurosci. 2022 Jan 31;16:760085. (PMID: 35173595) Cancers (Basel). 2019 May 24;11(5):. (PMID: 31137643) PLoS Comput Biol. 2023 Dec 11;19(12):e1011711. (PMID: 38079453) J Math Biol. 2023 Jul 11;87(2):27. (PMID: 37432484) |
| Entry Date(s): | Date Created: 20251106 Date Completed: 20251113 Latest Revision: 20251114 |
| Update Code: | 20251114 |
| PubMed Central ID: | PMC12611157 |
| DOI: | 10.1371/journal.pcbi.1013193 |
| PMID: | 41196942 |
| Databáza: | MEDLINE |
Full Text 
Full Text Finder 
Nájsť tento článok vo Web of Science | Abstrakt: | Competing Interests: The authors have declared that no competing interests exist.<br />Biological systems inherently exhibit multi-scale dynamics, making accurate system identification particularly challenging due to the complexity of capturing a wide time scale spectrum. Traditional methods capable of addressing this issue rely on explicit equations, limiting their applicability in cases where only observational data are available. To overcome this limitation, we propose a data-driven framework that integrates the Sparse Identification of Nonlinear Dynamics (SINDy) method, the multi scale analysis algorithm Computational Singular Perturbation (CSP) and neural networks (NNs). This framework allows the partition of the available dataset in subsets characterized by similar dynamics, so that system identification can proceed within these subsets without facing a wide time scale spectrum. Accordingly, when the full dataset does not allow SINDy to identify the proper model, CSP is employed for the generation of subsets of similar dynamics, which are then fed into SINDy. CSP requires the availability of the gradient of the vector field, which is estimated by the NNs. The framework is tested on the Michaelis-Menten model, for which various reduced models in analytic form exist at different parts of the phase space. It is demonstrated that the CSP-based data subsets allow SINDy to identify the proper reduced model in cases where the full dataset does not. In addition, it is demonstrated that the framework succeeds even in the cases where the available data set originates from stochastic versions of the Michaelis-Menten model. This framework is algorithmic, so system identification is not hindered by the dimensions of the dataset.<br /> (Copyright: © 2025 Muhammed et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.) |
|---|---|
| ISSN: | 1553-7358 |
| DOI: | 10.1371/journal.pcbi.1013193 |