Computational NeuroscienceThe MVGC multivariate Granger causality toolbox: A new approach to Granger-causal inference
Introduction
Wiener–Granger causality (G-causality) (Granger, 1969, Geweke, 1982, Geweke, 1984) is an increasingly popular method for identifying “causal” connectivity in neural time series data (Bressler and Seth, 2011). It can be traced conceptually to Wiener (1956) and was operationalised by Granger in terms of linear autoregressive modelling of stochastic processes (Granger, 1969). G-causality is based on predictability and precedence. Put simply, a variable X is said to G-cause a variable Y if the past of X contains information that helps predict the future of Y over and above information already in the past of Y. Importantly, G-causality is a measure of directed functional connectivity in terms of providing a statistical description of observed responses. In contrast, methods for identifying effective connectivity aim to elucidate the “the simplest possible circuit diagram explaining observed responses” (Valdes-Sosa et al., 2011, Aertsen and Preißl, 1991, Friston et al., 2003, Friston et al., 2013).
The MVGC Matlab Toolbox implements numerical routines for calculating multivariate Granger causality (MVGC) from time series data, both unconditional and conditional, in the time and frequency domains. It supersedes and extends the GCCA (Granger Causal Connectivity Analysis) Toolbox (Seth, 2010). Based on advanced VAR (vector autoregressive) model theory (Hamilton, 1994, Lütkepohl, 2005), it is designed for computational efficiency and accuracy. Improving upon standard approaches to G-causal inference, it avoids separate full and reduced regressions (see Section 2), thus eliminating a common source of statistical inaccuracy and computational inefficiency. It also facilitates estimation of fully multivariate conditional G-causality in the frequency domain.
The MVGC toolbox has been designed with application to empirical neuroscience data in mind. Several issues commonly faced in applying G-causal inference to such data are discussed later in the paper (Section 4). However, the software is not restricted to this application domain. G-causal inference is a very general method which has been productively applied in many areas, though caution is always needed in ensuring that the data satisfy the assumptions underpinning the method (see Section 2.1 and also Section 3.3). Thus, while the MVGC software is designed to be intuitive and straightforward to use, to get the most out of it some understanding of the theoretical basis of G-causal inference, as well as approaches to its numerical computation, is recommended. To facilitate this, this paper presents the conceptual and theoretical basis of G-causality (substantially extending previous presentations), and details the computational implementation of the MVGC toolbox. Although not intended as a user guide—the integrated help system and walk-through demonstration scripts fulfill that purpose—Section 3 provides a helpful overview of the toolbox rationale and design principles, Section 3.1 explains the MVGC computational strategy while A.1 Core algorithms, A.2 G-causality algorithms detail the core computational algorithms.
Section snippets
G-causality: theory, estimation and inference
Assume two jointly distributed vector-valued stochastic processes (“variables”) X = X1, X2, …, Y = Y1, Y2, …. We say that Y does not G-cause X if and only if X, conditional on its own past, is independent of the past of Y; intuitively, past values of Y yield no information about the current value of X beyond information already contained in the past of X itself. If, conversely, the past of Y does convey information about the future of X above and beyond all information contained in the past of X
MVGC Toolbox design
Central to the design of the MVGC toolbox is the equivalence (Section 2.1) of the VAR parameters (Ak, Σ), the autocovariance sequence Γk and the cross-power spectral density S(λ) as representations for a VAR process. The MVGC toolbox exploits these equivalences to provide numerically accurate algorithms for moving flexibly between the alternative VAR representations, thus furnishing computationally efficient pathways for calculation of G-causalities, conditional and unconditional, in the time
Application to neuroscience time series data
Although G-causality is an entirely general method for analysis of time series data, the MVGC toolbox has been developed with application to neuroscience data in mind. Methodological development in this application domain is rapidly advancing and a full review is beyond the present scope (see, for example, Deshpande and Hu, 2012, Friston et al., 2013, Bressler and Seth, 2011, Ding et al., 2006). This section summarises some of the main issues involved in application of G-causality (as
Conclusions
The MVGC toolbox provides a comprehensive set of Matlab routines for implementing G-causality analysis in the time and frequency domains and in both conditional and unconditional cases. It optimises computational efficiency, numerical accuracy and statistical inference by leveraging multiple equivalent representations of a VAR model by regression parameters, the autocovariance sequence, and the cross-power spectral density of the underlying process. By this approach, it is able to compute
Acknowledgements
We are grateful to Adam Barrett for helpful discussions and comments, and for testing pre-release versions of the MVGC toolbox. For financial support we are grateful to the EPSRC (G/700543/1) and to the Dr. Mortimer and Dame Theresa Sackler Foundation, which supports the work of the Sackler Centre for Consciousness Science.
Software availability: The MVGC toolbox may be downloaded from www.sussex.ac.uk/sackler. The software is freely available for non-profit academic usage under the GNU General
References (91)
- et al.
Behaviour of Granger causality under filtering: theoretical invariance and practical application
J Neurosci Methods
(2011) - et al.
Wiener–Granger causality: a well established methodology
Neuroimage
(2011) - et al.
Frequency decomposition of conditional Granger causality and application to multivariate neural field potential data
J Neurosci Methods
(2006) - et al.
Analyzing information flow in brain networks with nonparametric Granger causality
Neuroimage
(2008) - et al.
Influence analysis for high-dimensional time series with an application to epileptic seizure onset zone detection
J Neurosci Methods
(2013) - et al.
Analysing connectivity with Granger causality and dynamic causal modelling
Curr Opin Neurobiol
(2013) - et al.
Dynamic causal modelling
Neuroimage
(2003) Some properties of time series data and their use in econometric model specification
J Econometrics
(1981)- et al.
Partial granger causality: eliminating exogenous inputs and latent variables
J Neurosci Methods
(2008) - et al.
Variation of BOLD hemodynamic responses across subjects and brain regions and their effects on statistical analyses
Neuroimage
(2004)
Dynamic modeling of neuronal responses in fMRI using cubature Kalman filtering
Neuroimage
Information transfer in continuous processes
Physica D
Granger causality with signal-dependent noise
Neuroimage
Nonlinear denoising and analysis of neuroimages with kernel principal component analysis and pre-image estimation
Neuroimage
Mapping directed influence over the brain using granger causality and fMRI
Neuroimage
Does partial Granger causality really eliminate the influence of exogenous inputs and latent variables?
J Neurosci Methods
Multivariate dynamical systems models for estimating causal interactions in fMRI
Neuroimage
A MATLAB toolbox for Granger causal connectivity analysis
J Neurosci Methods
Granger causality analysis of fMRI BOLD signals is invariant to hemodynamic convolution but not downsampling
Neuroimage
Effective connectivity: influence, causality and biophysical modeling
Neuroimage
Estimating Granger causality after stimulus onset: a cautionary note
Neuroimage
A blind deconvolution approach to recover effective connectivity brain networks from resting state fMRI data
Med Image Anal
Dynamics of activity and connectivity in physiological neuronal networks
Permutation tests for linear models
Aust New Zeal J Stat
The statistical analysis of time series
Digital filters: analysis, design, and applications
Granger causality and transfer entropy are equivalent for Gaussian variables
Phys Rev Lett
Transfer entropy as a log-likelihood ratio
Phys Rev Lett
Multivariate Granger causality and generalized variance.
Phys Rev E
Granger causality analysis of steady-state electroencephalographic signals during propofol-induced anaesthesia
PLoS ONE
Solution of the equation AX + XB = C
Commun ACM
Causal measures of structure and plasticity in simulated and living neural networks
PLoS ONE
Stochastic modeling of neurobiological time series: power, coherence, Granger causality, and separation of evoked responses from ongoing activity
Chaos
Dynamic interactions between large-scale brain networks predict behavioral adaptation after perceptual errors
Cereb Cortex
An algorithm for the machine computation of the complex fourier series
Math Comput
Identifying neural drivers with functional MRI: an electrophysiological validation
PLoS Biol
Investigating effective brain connectivity from FMRI data: past findings and current issues with reference to granger causality analysis
Brain Connect
Estimating Granger causality from Fourier and wavelet transforms of time series data
Phys Rev Lett
Short-window spectral analysis of cortical event-related potentials by adaptive multivariate autoregressive modeling: data preprocessing, model validation, and variability assessment
Biol Cybern
Granger causality: basic theory and application to neuroscience
The fitting of time series models.
Rev Inst Int Stat
Likelihood (expanded edition)
The jackknife, the bootstrap, and other resampling plans
Society of Industrial and Applied Mathematics CBMS-NSF Monographs
Co-integration and error correction: representation, estimation and testing
Econometrica
The rapid development of high speed, resolution and precision in fMRI
Neuroimage
Cited by (682)
Graph attention network with Granger causality map for fault detection and root cause diagnosis
2024, Computers and Chemical EngineeringSynchronized LFP rhythmicity in the social brain reflects the context of social encounters
2024, Communications BiologyGranger causal representation learning for groups of time series
2024, Science China Information Sciences