Thus, there remain significant limitations on our ability to spatiotemporally resolve, either through direct observation or perturbationbased inference, how global information processing at the molecular scale gives rise to migratory behaviors. Given the modularity in molecular network structure noted above, an alternative strategy is to analyze the state and interactions of functional macromolecular modules within migrating cells. This provides a means to Yunaconitine coarse-grain the overwhelming molecular-scale complexity to a level that is tractable with imaging-based approaches, as recently demonstrated. However, even given such coarse-graining, it remains necessary to focus on a subset of selected modules that are central to the process of interest. In the case of cell migration, integrin-mediated cell-matrix adhesion complexes and the F-actin microfilament system are core macromolecular modules that directly mediate the migratory process. CMACs and F-actin serve as both mechanochemical-sensors and -transmitters, acting at the interface between cells and their environment to collect and functionalize information derived from extrinsic regulators in the distributed molecular network. As such, their status is innately linked to, and therefore partially representative of, the aggregate informational state of the cell migration system as a whole. Using live cell fluorescence imaging and quantitative image analyses, organizational features of the cell migration system can be extracted to create a quantitative, multivariate characterization of CMAC and F-actin status on a per cell basis, as well as of cellular scale morphology. Simultaneously, behavioral features can be recorded describing the migration of the same individual cells. This facilitates two of the critical enabling capabilities of the research framework described herein, Mycophenolic acid namely the: i) direct integration, and ii) temporal resolution of organizational and behavioral data on a per cell basis. Firstly, direct data integration on a per cell basis allows the covariance that arises between any two features as a result of natural cellular heterogeneity to define inter-feature correlations, making perturbations unnecessary to achieve this result. Crucially, this includes defining the relationships between features of cell organization and behavior. Secondly, the temporal resolution of variations in feature values allows us to move beyond correlative analyses to define the direction of information flow, in the form of causal influence, between pairs of system features using the Granger causality concept. Again, unlike many other 
systems-level causal analyses using, for example, network perturbation approaches, directed causal influences can be defined using Granger causality without the need for perturbations. The Granger causality concept originates from the field of econometrics, wherein the study of causality cannot routinely be facilitated by perturbation-based analysis. Similarly, Granger causality has become a key tool in the perturbationindependent mapping of neural connectivity and information processing. Herein we demonstrate one of the first implementations of this approach to the study of fundamental cell biology. Essentially, the Granger conception of causality stipulates causation if the combination of past information from both a background variable, X, and a response variable, Y, improves the prediction of future values of Y, when compared to prediction of Y based on only its own past values.