Synergy and redundancy in the Granger causal analysis of dynamical networks

Abstract

We analyze by means of Granger causality the effect of synergy and redundancy in the inference (from time series data) of the information flow between subsystems of a complex network. Whilst we show that fully conditioned Granger causality is not affected by synergy, the pairwise analysis fails to put in evidence synergetic effects. In cases when the number of samples is low, thus making the fully conditioned approach unfeasible, we show that partially conditioned Granger causality is an effective approach if the set of conditioning variables is properly chosen. We consider here two different strategies (based either on informational content for the candidate driver or on selecting the variables with highest pairwise influences) for partially conditioned Granger causality and show that depending on the data structure either one or the other might be valid. On the other hand, we observe that fully conditioned approaches do not work well in presence of redundancy, thus suggesting the strategy of separating the pairwise links in two subsets: those corresponding to indirect connections of the fully conditioned Granger causality (which should thus be excluded) and links that can be ascribed to redundancy effects and, together with the results from the fully connected approach, provide a better description of the causality pattern in presence of redundancy. We finally apply these methods to two different real datasets. First, analyzing electrophysiological data from an epileptic brain, we show that synergetic effects are dominant just before seizure occurrences. Second, our analysis applied to gene expression time series from HeLa culture shows that the underlying regulatory networks are characterized by both redundancy and synergy.

References in corpus (14)

• Measuring Information Transfer
• The large-scale organization of metabolic networks
• Functional cartography of complex metabolic networks
• Granger causality and transfer entropy are equivalent for Gaussian variables
• Nonlinear multivariate analysis of Neurophysiological Signals
• Network Physiology reveals relations between network topology and physiological function
• Kernel method for nonlinear Granger causality
• Multivariate Granger Causality and Generalized Variance
• A Bivariate Measure of Redundant Information
• Kernel Granger causality and the analysis of dynamical networks
• Expanding the Transfer Entropy to Identify Information Subgraphs in Complex Systems
• Towards a Synergy-based Approach to Measuring Information Modification
• Redundant variables and Granger causality
• Grouping time series by pairwise measures of redundancy

Cited by in corpus (16)

• Exploration of synergistic and redundant information sharing in static and dynamical Gaussian systems
• Bits from Biology for Computational Intelligence
• Multiscale Information Decomposition: Exact Computation for Multivariate Gaussian Processes
• Transfer entropy in continuous time, with applications to jump and neural spiking processes
• Quantifying information transfer and mediation along causal pathways in complex systems
• Inferring network properties from time series using transfer entropy and mutual information: validation of multivariate versus bivariate approaches
• Redundancy and synergy in dual decompositions of mutual information gain and information loss
• Causality in cognitive neuroscience: concepts, challenges, and distributional robustness
• Synergetic and redundant information flow detected by unnormalized Granger causality: application to resting state fMRI
• Generalised Measures of Multivariate Information Content
• Lagged and instantaneous dynamical influences related to brain structural connectivity
• A novel framework to analyze complex network dynamics
• A Graph Algorithmic Approach to Separate Direct from Indirect Neural Interactions
• Shortcomings of transfer entropy and partial transfer entropy: Extending them to escape the curse of dimensionality
• The identity of information: how deterministic dependencies constrain information synergy and redundancy
• Feedforward and feedback influences through distinct frequency bands between two spiking-neuron networks