Animated Spots & Stripes: Towards a Computational Developmental Science

During our interdisciplinary travels we have noted that biologists and ecologists often use methodological approaches that are, even with relatively small amounts of empirical data, informative about the latent processes underlying observed phenomena (i.e., the stuff under the blanket). Systems biology, for instance, makes use of computational approaches wherein theories of change are written out using a set of mathematical equations (i.e., the lingua franca of science), explicitly including parameters or variables that represent the specific mechanisms thought to be involved. So, we have started exploring the potential utility of reaction-diffusion system models in the understanding of developmental phenomena or processes like self-regulation. Using a Gierer-Meinhardt model we created another synthetic 100-year life span of data. Check it out.

How do we interpret what’s under that crazy blanket? We begin with random initial patterns in the activated (dark) regions locations, that may be considered as indicative of genetic sequences with long-lasting effects, as our other simulations show this initial pattern has an amount of control and influence over the resultant space. After the “birth” field is created synthetic development is shown in the animation, with many different patterns emerging as a function of the model kinetics; age-related change is here operationalized as change in rates of activation, inhibition, and diffusion within the space. And these changes capture processes of differentiation (e.g., development and emergence of spots), dedifferentiation (e.g., diffusion and flow from spots to stripe-like patterning), and redifferentiation (e.g., focused activation, return to a few spots) over our “life span” of time. While our social science interpretations remain somewhat fuzzy and speculative, the analogies to developmental phenomena such as earlier-life cognitive differentiation (e.g., the emerging spots) or self-identity, or patterns of behavior reinforcement; or later-life cognitive and physical decline, or increased rigidity in behavior or personality (e.g., as redifferentiation and decreased regions of activation) are seen as a few of potentially many interpretations of either the micro-time snapshots of pattern style or the macro-time view of pattern change. Having cracked open the door to a Computational Developmental Science based on such simulations – we are excited about a whole new set of possibilities.

In the mean time we try to get Robyn and Joey to make it into music!