scholarly journals A data-informed mean-field approach to mapping of cortical parameter landscapes

2021 ◽  
Vol 17 (12) ◽  
pp. e1009718
Author(s):  
Zhuo-Cheng Xiao ◽  
Kevin K. Lin ◽  
Lai-Sang Young
Keyword(s):  
Parameter Space ◽  
Large Scale ◽  
Mean Field ◽  
Network Models ◽  
Complete Characterization ◽  
Model Parameters ◽  
Cortical Model ◽  
Experimental Values ◽  
The Many ◽  
Data Informed

Constraining the many biological parameters that govern cortical dynamics is computationally and conceptually difficult because of the curse of dimensionality. This paper addresses these challenges by proposing (1) a novel data-informed mean-field (MF) approach to efficiently map the parameter space of network models; and (2) an organizing principle for studying parameter space that enables the extraction biologically meaningful relations from this high-dimensional data. We illustrate these ideas using a large-scale network model of the Macaque primary visual cortex. Of the 10-20 model parameters, we identify 7 that are especially poorly constrained, and use the MF algorithm in (1) to discover the firing rate contours in this 7D parameter cube. Defining a “biologically plausible” region to consist of parameters that exhibit spontaneous Excitatory and Inhibitory firing rates compatible with experimental values, we find that this region is a slightly thickened codimension-1 submanifold. An implication of this finding is that while plausible regimes depend sensitively on parameters, they are also robust and flexible provided one compensates appropriately when parameters are varied. Our organizing principle for conceptualizing parameter dependence is to focus on certain 2D parameter planes that govern lateral inhibition: Intersecting these planes with the biologically plausible region leads to very simple geometric structures which, when suitably scaled, have a universal character independent of where the intersections are taken. In addition to elucidating the geometry of the plausible region, this invariance suggests useful approximate scaling relations. Our study offers, for the first time, a complete characterization of the set of all biologically plausible parameters for a detailed cortical model, which has been out of reach due to the high dimensionality of parameter space.

scholarly journals A data-informed mean-field approach to mapping of cortical parameter landscapes

2021 ◽  
Author(s):  
Zhuo-Cheng Xiao ◽  
Kevin K Lin ◽  
Lai-Sang Young
Keyword(s):  
Parameter Space ◽  
Large Scale ◽  
Mean Field ◽  
Network Models ◽  
Complete Characterization ◽  
Model Parameters ◽  
Cortical Model ◽  
Experimental Values ◽  
The Many ◽  
Data Informed

Constraining the many biological parameters that govern cortical dynamics is computationally and conceptually difficult because of the curse of dimensionality. This paper addresses these challenges by proposing (1) a novel data-informed mean-field (MF) approach to efficiently map the parameter space of network models; and (2) an organizing principle for studying parameter space that enables the extraction biologically meaningful relations from this high-dimensional data. We illustrate these ideas using a large-scale network model of the Macaque primary visual cortex. Of the 10-20 model parameters, we identify 7 that are especially poorly constrained, and use the MF algorithm in (1) to discover the firing rate contours in this 7D parameter cube. Defining a "biologically plausible" region to consist of parameters that exhibit spontaneous Excitatory and Inhibitory firing rates compatible with experimental values, we find that this region is a slightly thickened codimension-1 submanifold. An implication of this finding is that while plausible regimes depend sensitively on parameters, they are also robust and flexible provided one compensates appropriately when parameters are varied. Our organizing principle for conceptualizing parameter dependence is to focus on certain 2D parameter planes that govern lateral inhibition: Intersecting these planes with the biologically plausible region leads to very simple geometric structures which, when suitably scaled, have a universal character independent of where the intersections are taken. In addition to elucidating the geometry of the plausible region, this invariance suggests useful approximate scaling relations. Our study offers, for the first time, a complete characterization of the set of all biologically plausible parameters for a detailed cortical model, which has been out of reach due to the high dimensionality of parameter space.

Canyon Building Ventilation System Dynamic Model Optimization Study

1993 ◽  
Author(s):  
Franklin F. K. Chen ◽  
B. Ronald Moncrief
Keyword(s):  
Dynamic Model ◽  
Parameter Space ◽  
Large Scale ◽  
Age Factors ◽  
Ventilation System ◽  
Operating Conditions ◽  
Data Availability ◽  
Model Parameters ◽  
Desktop Computer ◽  
Plant Data

Abstract A canyon building houses special nuclear material processing facilities in two canyon like structures, each with approximately a million cubic feet of air space and a hundred thousand hydraulic equivalent feet of ductwork of various cross sections. The canyon ventilation system is a “once through” design with separate supply and exhaust fans, utilizes two large sand filters to remove radionuclide particulate matter, and exhausts through a tall stack. The ventilation equipment is similar to most industrial ventilation systems. However, in a canyon building, nuclear contamination prohibits access to a large portion of the system and therefore limits the kind of plant data possible. The facility investigated is 40 years old and is operating with original or replacement equipment of comparable antiquity. These factors, access and aged equipment, present a challenge in gauging the performance of canyon ventilation, particularly under uncommon operating conditions. The ability to assess canyon ventilation system performance became critical with time, as the system took on additional exhaust loads and aging equipment approached design maximum. Many “What if?” questions, needed to address modernization/safety issues, are difficult to answer without a dynamic model. This paper describes the development, the validation and the utilization of a dynamic model to analyze the capacity of this ventilation system, under many unusual but likely conditions. The development of a ventilation model with volume and hydraulics of this scale is unique. The resultant model resolutions of better than 0.05″wg under normal plant conditions and approximately 0.2″wg under all plant conditions achievable with a desktop computer is a benchmark of the power of micro-computers. The detail planning and the persistent execution of large scale plant experiments under very restrictive conditions not only produced data to validate the model but lent credence to subsequent applications of the model to mission oriented analysis. Modelling methodology adopted a two parameter space approach, rational parameters and irrational parameters. Rational parameters, such as fan age-factors, idle parameters, infiltration areas and tunnel hydraulic parameters are deduced from plant data based on certain hydraulic models. Due to limited accessibility and therefore partial data availability, the identification of irrational model parameters, such as register positions and unidentifiable infiltrations, required unique treatment of the parameter space. These unique parameters were identified by a numerical search strategy to minimize a set of performance indices. With the large number of parameters, this further attests to our strategy in utilizing the computing power of modern micros. Nine irrational parameters at five levels and 12 sets of plant data, counting up to 540 runs, were completely searched over the time span of a long weekend. Some key results, in assessing emergency operation, in evaluating modernization options, are presented to illustrate the functions of the dynamic model.

scholarly journals Cluster mean-field theory accurately predicts statistical properties of large-scale DNA methylation patterns

2021 ◽  
Author(s):  
Lyndsay Kerr ◽  
Duncan Sproul ◽  
Ramon Grima
Keyword(s):  
Dna Methylation ◽  
Large Scale ◽  
Field Model ◽  
Mean Field ◽  
Statistical Properties ◽  
Stochastic Simulations ◽  
Model Parameters ◽  
Mean Field Model ◽  
Mean Field Models ◽  
Methylation Patterns

The accurate establishment and maintenance of DNA methylation patterns is vital for mammalian development and disruption to these processes causes human disease. Our understanding of DNA methylation mechanisms has been facilitated by mathematical modelling, particularly stochastic simulations. Mega-base scale variation in DNA methylation patterns is observed in development, cancer and ageing and the mechanisms generating these patterns are little understood. However, the computational cost of stochastic simulations prevents them from modelling such large genomic regions. Here we test the utility of three different mean-field models to predict large-scale DNA methylation patterns. By comparison to stochastic simulations, we show that a cluster mean-field model accurately predicts the statistical properties of steady-state DNA methylation patterns, including the mean and variance of methylation levels calculated across a system of CpG sites, as well as the covariance and correlation of methylation levels between neighbouring sites. We also demonstrate that a cluster mean-field model can be used within an approximate Bayesian computation framework to accurately infer model parameters from data. As mean-field models can be solved numerically in a few seconds, our work demonstrates their utility for understanding the processes underpinning large-scale DNA methylation patterns.

Adjoint Sensitivity Analysis of a Cooling Tower

2017 ◽  
Author(s):  
Federico Di Rocco ◽  
Dan Gabriel Cacuci
Keyword(s):  
Nuclear Power ◽  
Power Plants ◽  
Large Scale ◽  
External Environment ◽  
Model Parameters ◽  
Adjoint Sensitivity ◽  
Analysis Methodology ◽  
Functional Derivatives ◽  
Adjoint Analysis ◽  
The Many

Every plant aimed for energy production on large scale mandatorily needs to include systems able to ensure a proper release of the residual heat into the external environment; in the case of nuclear power plants, cooling towers are the most popular answer to this compelling task. The development of a dedicated numerical model can be a powerful tool for the simulation of the tower performance, but for quantifying the amount of heat released into the atmosphere, experimentally measured data are needed, in order to properly characterize the external environment in terms of outlet air temperature, outlet water temperature and outlet air relative humidity. Material properties, correlations and boundary conditions are among the many model parameters which can influence the model’s responses; variations of the computed quantities of interest (or “model responses”) induced by variations in the values of the model parameters can be expressed in terms of the sensitivities (i.e., functional derivatives) of the model responses with respect to the model parameters. The methodology applied in this paper in order to compute these sensitivities is the general adjoint analysis methodology (ASAM) for nonlinear systems; with this procedure, the exact values of the sensitivities of the model responses to all of the 47 model parameters are calculated exactly and efficiently by means of a single adjoint computation. The sensitivities can then be used within the “predictive modeling for coupled multi-physics systems” (PM_CMPS) methodology, aimed at yielding best-estimate predicted nominal values and uncertainties for model parameters and responses.

scholarly journals Biophysically motivated regulatory network inference: progress and prospects

2016 ◽  
Author(s):  
Tarmo Äijö ◽  
Richard Bonneau
Keyword(s):  
Regulatory Network ◽  
Large Scale ◽  
Network Inference ◽  
Network Models ◽  
Cell Types ◽  
Parameters Estimation ◽  
Model Parameters ◽  
Regulatory Factor ◽  
Regulatory Structure ◽  
Dynamic Genome

AbstractVia a confluence of genomic technology and computational developments the possibility of network inference methods that automatically learn large comprehensive models of cellular regulation is closer than ever. This perspective will focus on enumerating the elements of computational strategies that, when coupled to appropriate experimental designs, can lead to accurate large-scale models of chromatin-state and transcriptional regulatory structure and dynamics. We highlight four research questions that require further investigation in order to make progress in network inference: using overall constraints on network structure like sparsity, use of informative priors and data integration to constrain individual model parameters, estimation of latent regulatory factor activity under varying cell conditions, and new methods for learning and modeling regulatory factor interactions. We conclude that methods combining advances in these four categories of required effort with new genomic technologies will result in biophysically motivated dynamic genome-wide regulatory network models for several of the best studied organisms and cell types.

scholarly journals Simulating large-scale models of brain neuronal circuits using Google Cloud Platform

2020 ◽  
Author(s):  
Subhashini Sivagnanam ◽  
Wyatt Gorman ◽  
Donald Doherty ◽  
Samuel A Neymotin ◽  
Stephen Fang ◽  
...  
Keyword(s):  
Large Scale ◽  
Circuit Simulation ◽  
Experimental Studies ◽  
Model Parameters ◽  
Cortical Model ◽  
Detailed Model ◽  
Scale Models ◽  
Brain Circuit ◽  
Large Scale Simulations ◽  
The Brain

Biophysically detailed modeling provides an unmatched method to integrate data from many disparate experimental studies, and manipulate and explore with high precision the resulting brain circuit simulation. We developed a detailed model of the brain motor cortex circuits, simulating over 10,000 biophysically detailed neurons and 30 million synaptic connections. Optimization and evaluation of the cortical model parameters and responses was achieved via parameter exploration using grid search parameter sweeps and evolutionary algorithms. This involves running tens of thousands of simulations, with each simulated second of the full circuit model requiring approximately 50 cores hours. This paper describes our experience in setting up and using Google Compute Platform (GCP) with Slurm to run these large-scale simulations. We describe the best practices and solutions to the issues that arose during the process, and present preliminary results from running simulations on GCP.

scholarly journals Optimal responsiveness and information flow in networks of heterogeneous neurons

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Matteo Di Volo ◽  
Alain Destexhe
Keyword(s):  
Information Flow ◽  
Heterogeneous Networks ◽  
Large Scale ◽  
Computational Models ◽  
Field Model ◽  
Mean Field ◽  
Network Models ◽  
Inhibitory Neurons ◽  
Mean Field Model ◽  
Optimal Responsiveness

AbstractCerebral cortex is characterized by a strong neuron-to-neuron heterogeneity, but it is unclear what consequences this may have for cortical computations, while most computational models consider networks of identical units. Here, we study network models of spiking neurons endowed with heterogeneity, that we treat independently for excitatory and inhibitory neurons. We find that heterogeneous networks are generally more responsive, with an optimal responsiveness occurring for levels of heterogeneity found experimentally in different published datasets, for both excitatory and inhibitory neurons. To investigate the underlying mechanisms, we introduce a mean-field model of heterogeneous networks. This mean-field model captures optimal responsiveness and suggests that it is related to the stability of the spontaneous asynchronous state. The mean-field model also predicts that new dynamical states can emerge from heterogeneity, a prediction which is confirmed by network simulations. Finally we show that heterogeneous networks maximise the information flow in large-scale networks, through recurrent connections. We conclude that neuronal heterogeneity confers different responsiveness to neural networks, which should be taken into account to investigate their information processing capabilities.

scholarly journals Inhibitory control of correlated intrinsic variability in cortical networks

2016 ◽  
Author(s):  
Carsen Stringer ◽  
Marius Pachitariu ◽  
Michael Okun ◽  
Peter Bartho ◽  
Kenneth Harris ◽  
...  
Keyword(s):  
Large Scale ◽  
Feedback Inhibition ◽  
Activity Patterns ◽  
Network Models ◽  
External Noise ◽  
Inhibitory Neurons ◽  
Computational Techniques ◽  
Model Parameters ◽  
Intrinsic Variability ◽  
Cortical Networks

AbstractCortical networks exhibit intrinsic dynamics that drive coordinated, large-scale fluctuations across neuronal populations and create noise correlations that impact sensory coding. To investigate the network-level mechanisms that underlie these dynamics, we developed novel computational techniques to fit a deterministic spiking network model directly to multi-neuron recordings from different species, sensory modalities, and behavioral states. The model generated correlated variability without external noise and accurately reproduced the wide variety of activity patterns in our recordings. Analysis of the model parameters suggested that differences in noise correlations across recordings were due primarily to differences in the strength of feedback inhibition. Further analysis of our recordings confirmed that putative inhibitory neurons were indeed more active during desynchronized cortical states with weak noise correlations. Our results demonstrate that network models with intrinsically-generated variability can accurately reproduce the activity patterns observed in multi-neuron recordings and suggest that inhibition modulates the interactions between intrinsic dynamics and sensory inputs to control the strength of noise correlations.

scholarly journals Inhibitory control of correlated intrinsic variability in cortical networks

eLife ◽  
2016 ◽  
Vol 5 ◽  
Author(s):  
Carsen Stringer ◽  
Marius Pachitariu ◽  
Nicholas A Steinmetz ◽  
Michael Okun ◽  
Peter Bartho ◽  
...  
Keyword(s):  
Large Scale ◽  
Feedback Inhibition ◽  
Activity Patterns ◽  
Network Models ◽  
External Noise ◽  
Inhibitory Neurons ◽  
Computational Techniques ◽  
Model Parameters ◽  
Intrinsic Variability ◽  
Cortical Networks

Cortical networks exhibit intrinsic dynamics that drive coordinated, large-scale fluctuations across neuronal populations and create noise correlations that impact sensory coding. To investigate the network-level mechanisms that underlie these dynamics, we developed novel computational techniques to fit a deterministic spiking network model directly to multi-neuron recordings from different rodent species, sensory modalities, and behavioral states. The model generated correlated variability without external noise and accurately reproduced the diverse activity patterns in our recordings. Analysis of the model parameters suggested that differences in noise correlations across recordings were due primarily to differences in the strength of feedback inhibition. Further analysis of our recordings confirmed that putative inhibitory neurons were indeed more active during desynchronized cortical states with weak noise correlations. Our results demonstrate that network models with intrinsically-generated variability can accurately reproduce the activity patterns observed in multi-neuron recordings and suggest that inhibition modulates the interactions between intrinsic dynamics and sensory inputs to control the strength of noise correlations.

scholarly journals Brain-scale emergence of slow-wave synchrony and highly responsive asynchronous states based on biologically realistic population models simulated in The Virtual Brain

2020 ◽  
Author(s):  
Jennifer S. Goldman ◽  
Lionel Kusch ◽  
Bahar Hazal Yalcinkaya ◽  
Damien Depannemaecker ◽  
Trang-Anh E. Nghiem ◽  
...  
Keyword(s):  
Slow Wave ◽  
Large Scale ◽  
Slow Wave Sleep ◽  
Mean Field ◽  
Population Models ◽  
Slow Waves ◽  
Brain Dynamics ◽  
Promising Tool ◽  
Activity State ◽  
The Many

ABSTRACTUnderstanding the many facets of the organization of brain dynamics at large scales remains largely unexplored. Here, we construct a brain-wide model based on recent progress in biologically-realistic population models obtained using mean-field techniques. We use The Virtual Brain (TVB) as a simulation platform and incorporate mean-field models of networks of Adaptive Exponential (AdEx) integrate-and-fire neurons. Such models can capture the main intrinsic firing properties of central neurons, such as adaptation, and also include the typical kinetics of postsynaptic conductances. We hypothesize that such features are important to a biologically realistic simulation of brain dynamics. The resulting “TVB-AdEx” model is shown here to generate two fundamental dynamical states, asynchronous-irregular (AI) and Up-Down states, which correspond to the asynchronous and synchronized dynamics of wakefulness and slow-wave sleep, respectively. The synchrony of slow waves appear as an emergent property at large scales, and reproduce the very different patterns of functional connectivity found in slow-waves compared to asynchronous states. Next, we simulated experiments with transcranial magnetic stimulation (TMS) during asynchronous and slow-wave states, and show that, like in experimental data, the effect of the stimulation greatly depends on the activity state. During slow waves, the response is strong but remains local, in contrast with asynchronous states, where the response is weaker but propagates across brain areas. To compare more quantitatively with wake and slow-wave sleep states, we compute the perturbational complexity index and show that it matches the value estimated from TMS experiments. We conclude that the TVB-AdEx model replicates some of the properties of synchrony and responsiveness seen in the human brain, and is a promising tool to study spontaneous and evoked large-scale dynamics in the normal, anesthetized or pathological brain.

Sign in / Sign up

Export Citation Format

Share Document