Feedback Control in Biology

Feedback Control in Biology

Calcium levels of pregnant cows in Iowa may not be the first image that comes to mind when Christian apologists invoke supernatural design, but that appears to be the case based on a recent feedback control study. As an electrical engineer, I can appreciate the level of design and fine-tuning required in making such systems work properly. The next several paragraphs examine and compare (in some technical detail) human-designed control systems with natural ones found in cows.

In engineering, feedback control is a common method used to maintain specified levels of important system outputs and quantities in the face of a variety of disturbances. A very typical design approach used to accomplish this is a 2nd order control system using a Proportional plus Integral (PI) controller. In this scheme, the overall system, the PI controller, and the two associated controller constants1 are very carefully designed to produce the desired transient system response. Such a response seeks an appropriate trade-off between the settling time, overshoot, and oscillatory behavior, to name a few characteristics of interest. An optimally designed control system will ensure that the system output tracks to the desired level after a disturbance within a reasonable time and with a reasonable transient characteristic.

(See here for a schematic diagram of a typical feedback control system.)

Apparently, God is an engineer, and a very good one. The very same scheme of feedback control as described above and the network topology reflected in the diagram has been found in mammals for calcium homeostasis, which is the regulation of plasma calcium concentration (calcium concentration in the blood). (See hereand here for the researchers’ reports.)

In the biological world, calcium homeostasis is necessary for the survival of mammals. The plasma calcium concentration needs to be maintained very tightly in mammals2 in spite of various disturbances related to diet or the calcium demand to meet milk production and fetal growth needs. Researchers studied the transient response of calcium concentration of a total of 38 dairy cows during a 10-day period surrounding the process of calving. They concluded that the cows were able to maintain a life-essential level of calcium in the blood only with the aid of the functional equivalent of a 2nd order PI feedback control system. As their studies show, such a control system can be realistically implemented biologically using two hormones. This biological control system is a strict and rigorous analogy to 2nd order PI control systems widely used in engineering applications.

For one engineering example, I have recently completed the design and characterization of a 2nd order PI control system for a communications application. The two controller constants were judiciously calculated, tested, and the final set chosen to produce the best transient response, thereby optimizing the overall system performance. A scaled response of this control system was compared with the calcium concentration experimental data 3 and was found to match very closely. 4 This suggests to me that in addition to the design implied by having the correct network topology, the two effective constants for the biological controller have likewise been judiciously “chosen” to produce the best transient response and, thereby, provide optimal results for the mammals subject to stringent calcium concentration regulatory demands.

This analogy between manmade and natural control systems implies stringent design at various levels, irreducible complexity, and fine-tuning. The researchers seem to recognize the level of complexity involved.

Yet, the most important implication of integral feedback does not lie in producing a simple dynamical model that agrees well with the actual data. Rather, it lies in the severe structural constraints that it imposes in the underlying homeostatic mechanism.

The “severe structural constraints” associated with the homeostatic mechanism, coupled with the seemingly optimal controller constants, display a system that has been designed and fine-tuned.

The researchers also reference the results of other research in conjunction with their own.

These results as well as those reported in this article seem to point to the prevalence of integral control in mechanisms where physiological quantities must be maintained with a narrow range despite internal and external disturbances.

Finally, they suggest that “[f]urther work is needed to catalog and uncover the architecture of these systems where integral control is at work.”

This analogy specifically, and feedback control in biological systems in general, reinvigorates William Paley’s famous Watchmaker argument. The biological feedback control system discussed in this article is the direct counterpart to feedback control systems used in engineering, systems that without question are recognized to require design and exquisite fine-tuning by intelligent agents. Consistent logic suggests that a divine Intelligent Agent is responsible for similar systems found in biology.

Such purposeful design and fine-tuning fits nicely with Christian theism and with the Reasons To Believe creation model.


Keith McPherson

Keith McPherson received his Master of Science in Electrical Engineering from Georgia Institute of Technology in 1993, and currently works as an electrical engineer in Melbourne, FL, in the fields of communications and signal processing.


Endnotes
  1. For a typical 2nd order PI control system, the designer has two constants to specify as part of the system design: the proportional constant (Kp), and the integral constant (Ki).

  2. 0.085-0.105 g/l in humans and 0.08-0.1 g/l in dairy cows.

  3. See Figure 2.b for the biological transient response.

  4. In addition to the characteristic shape of the two responses, the damping ratio was investigated. The so-called damping ratio is a parameter used to design and characterize control systems in engineering. The mentioned engineering control system was designed for a damping ratio of exactly 1.0. Analysis and side by side comparison of the two transient responses suggest that the damping ratio of the calcium homeostasis PI control system is slightly larger than 1.0, but probably no larger than 1.1. Damping ratios on this order are known to yield good system results in many engineering contexts. In engineering applications, the damping ratio typically lies between 0.5 and 2. A system with a damping ratio of 1.0 is referred to as a critically damped system. These systems converge faster than any other without oscillating. See here for more information on damping.