Exploring the multiscale world of biological dynamics: From concepts to computational tools

Stephen Hawking has called our times the "century of complexity." The more scientists probe the physical world that lies outside of us, the more complex it appears. Such complexity reveals itself in the unpredictable changes in weather patterns, the swirls of planetary gases and even the bubbling of soup on the stove. But likely even more complex than the physical world around us is the world within us, namely the realm of physiology.

Complexity, whether in the physical or physiologic worlds, is fundamentally different from "complicatedness." In everyday language, the terms 'complex' and 'complicated' are often used interchangeably. But, scientifically, they refer to very different types of structures and dynamics. As an example, modern machines can be exceptionally complicated, even with the user's manual. However, the flocking behavior of birds, the music of Mozart, and even the dynamics of the human heartbeat observed at rest or during sleep, are inherently complex. Furthermore, not all types of variability are equally complex. My colleagues and I have shown that the erratic changes in the heart's beating associated with a very common cardiac rhythm disturbance called atrial fibrillation are, paradoxically, less complex than the much more subtle heartbeat fluctuations associated with normal cardiac activation.

What are the properties that make some systems complex and not just complicated? Trying to answer this question is a major motivation underlying the exploration of biologic systems proposed here, and indeed, is the propelling force in my career as a physicist working at the interdisciplinary interfaces of basic science and biomedicine.

A key feature of healthy biologic systems is their capacity to adapt - to tune their dynamics and even re-sculpt their anatomies to cope with environments that are usually unpredictable and often dangerous or destructive. The marvelous plasticity of the human brain after stroke is one example of a recuperative type of adaptive capacity. The ability of world-class skiers to navigate the turns and twists of the slalom course at extreme speed is another. The term, and the foundational conceptual framework, of complex adaptive systems, was proposed and developed by Professor John H. Holland at the University of Michigan and his colleagues. This linkage between adaptability and complexity underlies my efforts to probe systems in health and their breakdown with a variety of pathologies.

From a dynamical viewpoint, the properties that enable this type of creative adaptability have esoteric names that make them seem unusual and even abnormal. These properties include nonlinearity (things do not add up), nonstationarity (things do not sit still), the lack of a single or characteristic scale (one finds fractal or tree-like organization) and non-equilibrium (statistically improbable) dynamics and spatial patterns. Yet, without these negative sounding features, life would not be possible.

Translated into more accessible language, nonlinearity refers to systems that do not respond in a way that is proportional to the amount they are stimulated or perturbed. Small changes can have huge effects ("the straw that broke the camels back") and conversely, large perturbations may induce only minor changes. Nonlinearity, however, does not imply anarchy; nonlinear systems do have organizing, discoverable principles. The challenge, reflecting the initial quote from Stephen Hawking, is that these principles cannot be uncovered by the traditional reductionist strategy in which scientists dissect a system into its constituent pieces, study each component in detail, and finally put them back together in an attempt to recreate the original entity. Reductionism, while essential to science, is not sufficient to provide a full understanding of systems that are nonlinear.

In everyday parlance, nonlinearity is responsible for the familiar notion of the whole being greater than the sum of the parts. In the technical language of complex systems, it is known by the term emergent properties. For nonlinear systems, the composite or group behavior (of molecules, cells, organs, flocks of birds and human societies) cannot be fully understood by simply adding up the components. Nonlinearity requires rigorous, new approaches to measuring and modeling the behavior of both mathematical and "real-world" living systems.

A relatively recent development central to this proposal has been the reexamination of system dynamics in the context of health and disease. Not unexpectedly, healthy and pathologic systems have very different dynamical patterns. Particularly compelling from my perspective is the finding that certain features appear to be common, and perhaps even universal, to the dynamics of health and disease, respectively. Further, quantitative analysis of systems that have become dysfunctional reveals certain themes (dynamical signatures) that emerge in contexts as seemingly different as those regulating the heartbeat and the immune system.

A continuing source of surprise to me in my own investigations, which scale up from red blood cell motions on the microscopic level to human balance studies on the macroscopic level, is the presence of 1) similar patterns of complexity in different types of healthy, adaptive systems, and 2) the breakdown or even frank collapse of complexity with disease and biotoxicity. Whether these physiologic principles apply at even larger scales, namely at societal and ecosystem levels, is an intriguing possibility.

On a practical level, more basic understanding and advanced computational tools point the way to new ways to monitor health status, measure the aging process, detect drug toxicity and forecast catastrophic events such as seizures and sudden cardiac death, as well as the need for life-saving interventions in trauma victims. These computational tools also hold the promise to provide a new class of "dynamical assays" to rapidly screen and test new therapeutic interventions designed to restore complexity and enhance system adaptiveness.