Behavioral plasticity may arise mechanistically from two distinct types of variation within the underlying network: 1) variation in the properties of features within the network, and/or; 2) variations in the connectivity between those features. To date, numerous studies have described how specific perturbations induce Lithium citrate changes in the values of features-of-interest within the cell migration system. In notable examples, extensive arrays of F-actin, CMAC and cell behavioral properties were monitored and compared across a spectrum of conditions wherein ECM-density, intracellular contractility and/or growth factor stimulus were modulated. Although important correlative links were implied by these studies, direct quantitative correlations between many of the recorded features were inaccessible because data were derived from distinct experimental sources. This highlights the advantages of defining inter-feature connectivity based on multivariate data derived from within individual experimental sources and, indeed, from within individual conditions. Such strategies rely on sensitivity to natural heterogeneity within conditions, thereby allowing the detection of variations in the inter-feature connectivity between conditions. A highly relevant study where this was MDL-29951 achieved includes an analysis of how correlative connections between CMAC morphology-features were altered by individual siRNAs within an RNAi screen. Indeed, this study is particularly noteworthy because it documents plasticity in the correlative relationships between morphological CMAC features. Our work now builds on elements of this important foundation in two ways: 1) assessment of the relationships between features of cell migration system organization and behavior, as facilitated by synchronous acquisition of both data types on a per cell basis, and; 2) definition of the direction of causal influences between features enabled by time-resolved data and Granger causality analysis. The extension of correlative analyses to delineate causal influence marks a significant advance. As noted in a recent commentary, the majority of analyses in cell biology are correlative, with the analysis of causation remaining infrequent. This is particularly true at a systemic level where complex patterns of inter-feature causation are considered. However, one prominent strategy for empirically defining causal wiring patterns is network perturbation. Generally this approach involves the characterization of an array of features in both the presence and absence of perturbation. The variation in feature values enforced by perturbations is then used to detect connectivity between feature-pairs and, specifically, inter-feature causation. Although extremely powerful, some limitations to the network perturbation strategy may bear further consideration here. Firstly, as with more locally focused epistasis strategies, inter-feature connectivity is defined by comparison of feature values between control and perturbed conditions. Because the combination of conditions is used to define a single causal wiring pattern, network plasticity cannot easily be addressed through the comparison of wiring patterns. A noteworthy exception to this is the timeresolved network perturbation analysis performed by Ku et al, which demonstrated the temporal evolution of causal influence patterns between coarse-grained features of microtubule, F-actin and myosin machineries during neutrophil polarization.