Multiple model-informed open-loop control of uncertain intracellular signaling dynamics.
Computational approaches to tune the activation of intracellular signal transduction pathways both predictably and selectively will enable researchers to explore and interrogate cell biology with unprecedented precision. Techniques to control complex nonlinear systems typically involve the applicati...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2014-04-01
|
| Series: | PLoS Computational Biology |
| Online Access: | http://europepmc.org/articles/PMC3983080?pdf=render |
| _version_ | 1819091079834632192 |
|---|---|
| author | Jeffrey P Perley Judith Mikolajczak Marietta L Harrison Gregery T Buzzard Ann E Rundell |
| author_facet | Jeffrey P Perley Judith Mikolajczak Marietta L Harrison Gregery T Buzzard Ann E Rundell |
| author_sort | Jeffrey P Perley |
| collection | DOAJ |
| description | Computational approaches to tune the activation of intracellular signal transduction pathways both predictably and selectively will enable researchers to explore and interrogate cell biology with unprecedented precision. Techniques to control complex nonlinear systems typically involve the application of control theory to a descriptive mathematical model. For cellular processes, however, measurement assays tend to be too time consuming for real-time feedback control and models offer rough approximations of the biological reality, thus limiting their utility when considered in isolation. We overcome these problems by combining nonlinear model predictive control with a novel adaptive weighting algorithm that blends predictions from multiple models to derive a compromise open-loop control sequence. The proposed strategy uses weight maps to inform the controller of the tendency for models to differ in their ability to accurately reproduce the system dynamics under different experimental perturbations (i.e. control inputs). These maps, which characterize the changing model likelihoods over the admissible control input space, are constructed using preexisting experimental data and used to produce a model-based open-loop control framework. In effect, the proposed method designs a sequence of control inputs that force the signaling dynamics along a predefined temporal response without measurement feedback while mitigating the effects of model uncertainty. We demonstrate this technique on the well-known Erk/MAPK signaling pathway in T cells. In silico assessment demonstrates that this approach successfully reduces target tracking error by 52% or better when compared with single model-based controllers and non-adaptive multiple model-based controllers. In vitro implementation of the proposed approach in Jurkat cells confirms a 63% reduction in tracking error when compared with the best of the single-model controllers. This study provides an experimentally-corroborated control methodology that utilizes the knowledge encoded within multiple mathematical models of intracellular signaling to design control inputs that effectively direct cell behavior in open-loop. |
| first_indexed | 2024-12-21T22:34:02Z |
| format | Article |
| id | doaj.art-ad9bf01d162e485fb5f266261066d568 |
| institution | Directory Open Access Journal |
| issn | 1553-734X 1553-7358 |
| language | English |
| last_indexed | 2024-12-21T22:34:02Z |
| publishDate | 2014-04-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS Computational Biology |
| spelling | doaj.art-ad9bf01d162e485fb5f266261066d5682022-12-21T18:48:01ZengPublic Library of Science (PLoS)PLoS Computational Biology1553-734X1553-73582014-04-01104e100354610.1371/journal.pcbi.1003546Multiple model-informed open-loop control of uncertain intracellular signaling dynamics.Jeffrey P PerleyJudith MikolajczakMarietta L HarrisonGregery T BuzzardAnn E RundellComputational approaches to tune the activation of intracellular signal transduction pathways both predictably and selectively will enable researchers to explore and interrogate cell biology with unprecedented precision. Techniques to control complex nonlinear systems typically involve the application of control theory to a descriptive mathematical model. For cellular processes, however, measurement assays tend to be too time consuming for real-time feedback control and models offer rough approximations of the biological reality, thus limiting their utility when considered in isolation. We overcome these problems by combining nonlinear model predictive control with a novel adaptive weighting algorithm that blends predictions from multiple models to derive a compromise open-loop control sequence. The proposed strategy uses weight maps to inform the controller of the tendency for models to differ in their ability to accurately reproduce the system dynamics under different experimental perturbations (i.e. control inputs). These maps, which characterize the changing model likelihoods over the admissible control input space, are constructed using preexisting experimental data and used to produce a model-based open-loop control framework. In effect, the proposed method designs a sequence of control inputs that force the signaling dynamics along a predefined temporal response without measurement feedback while mitigating the effects of model uncertainty. We demonstrate this technique on the well-known Erk/MAPK signaling pathway in T cells. In silico assessment demonstrates that this approach successfully reduces target tracking error by 52% or better when compared with single model-based controllers and non-adaptive multiple model-based controllers. In vitro implementation of the proposed approach in Jurkat cells confirms a 63% reduction in tracking error when compared with the best of the single-model controllers. This study provides an experimentally-corroborated control methodology that utilizes the knowledge encoded within multiple mathematical models of intracellular signaling to design control inputs that effectively direct cell behavior in open-loop.http://europepmc.org/articles/PMC3983080?pdf=render |
| spellingShingle | Jeffrey P Perley Judith Mikolajczak Marietta L Harrison Gregery T Buzzard Ann E Rundell Multiple model-informed open-loop control of uncertain intracellular signaling dynamics. PLoS Computational Biology |
| title | Multiple model-informed open-loop control of uncertain intracellular signaling dynamics. |
| title_full | Multiple model-informed open-loop control of uncertain intracellular signaling dynamics. |
| title_fullStr | Multiple model-informed open-loop control of uncertain intracellular signaling dynamics. |
| title_full_unstemmed | Multiple model-informed open-loop control of uncertain intracellular signaling dynamics. |
| title_short | Multiple model-informed open-loop control of uncertain intracellular signaling dynamics. |
| title_sort | multiple model informed open loop control of uncertain intracellular signaling dynamics |
| url | http://europepmc.org/articles/PMC3983080?pdf=render |
| work_keys_str_mv | AT jeffreypperley multiplemodelinformedopenloopcontrolofuncertainintracellularsignalingdynamics AT judithmikolajczak multiplemodelinformedopenloopcontrolofuncertainintracellularsignalingdynamics AT mariettalharrison multiplemodelinformedopenloopcontrolofuncertainintracellularsignalingdynamics AT gregerytbuzzard multiplemodelinformedopenloopcontrolofuncertainintracellularsignalingdynamics AT annerundell multiplemodelinformedopenloopcontrolofuncertainintracellularsignalingdynamics |