Biological organisation can be understood as an interlinked system of closed-loop control systems which can be characterised independently of the physical composition of the organs and molecules of which they are comprised.
These systems are well studied in the areas of engineering and cybernetics but are equally applicable to biological systems where similar patterns emerge. The closed loop control system is a specific type of network system (as studied by Bio-regulatory medicine) with an added feedback loop which gives the whole system purpose and stability to perturbations via complex and quasi-intelligent behaviour.
Control systems are essential to the understanding of:
- Homeostasis as a definition of health
- The understanding of disease as dis-regulation
- A correct assessment of dose-response relationships
- Rhythmic behaviour of biological systems
- The phenomenon of goal-oriented behaviour
- The solution to the problem of stability in the presence of disturbances
Control systems are portrayed in slightly different ways but all contain the same basic elements.
- Input to the system (eg. heat from sunlight)
- The system – responsible for performing some biological function (cooling down)
- Disturbance or noise input into the system – (exercise)
- Output – a desired result from the process (correct body temperature)
- Feedback from the output back into the system (controller)
It is this last item, feedback, that gives the system its complex characteristics and makes it ideal for biological regulation. In temperature regulation for example, we can imagine that the body somehow monitors all the heat coming in from the outside and calculates all the heat generated by the body to work out how much energy needs to be lost to maintain a constant temperature. This is just not practical however as small errors will be made which could accumulate over time and result in over-heating and death,
The way to proceed then is for the body to monitor its own temperature and adjust heat loss accordingly.
Biological Control Systems: Systems Biology of Diseases and the Design of Effective Treatments – Babatunde Ogunnaike – paper
“Control is everywhere in biological systems; in fact, physiological life as we know it is not possible without control“
Feedback control systems exist at all scales of biological organisation:
- System wide regulation – nervous, cardiovascular systems..
- Cellular processes – growth regulation, cellular division, DNA repair..
- Molecular processes – gene expression, protein synthesis, metabolite regulation..
Biological processes are obviously complex but if they can be characterised as a control system then things are simplified and the system can be studied according to already existing principles.
Disease, from this perspective, is a malfunctioning control system and depending upon which part of the system has gone wrong, typical patterns of malfunction will be observed.
“A short term but significant departure of physiological variables from normal limits is referred to as an ‘illness’ where recovery occurs when the perturbed variables are
returned to within normal limits. ‘Chronic illnesses’ are characterized by long term departures from normal limits, but with the affected variables still within nonlethal limits. Death occurs when critical physiological variables fall outside non-lethal limits and are not restored within a reasonable period of time.” – Ogunnaike
An example of diabetes is given whereby the two types of disease are caused by malfunction in two different parts of the control system:
- Type 1 diabetes: Non-functioning controller (pancreas not producing insulin)
- Type 2 diabetes: Faulty sensor (insulin resistance) leading to bad feedback
Diagnosis is by a perturbation of the system (administration of glucose) and examination of the output (blood glucose levels) for a characteristic response. Knowledge of the bio-chemistry here is irrelevant as the dose response relationship gives sufficient information for diagnosis.
Platelet count control. The chart shows the dose-response curve when a patient with a low platelet count is administered a dose of romiplostim. The platelet counts increases as desired but, crucially, only after a time delay of about 14 days.
The requirement is for a stable elevated platelet count but what happens in practice is that doses are administered every two weeks and an oscillating platelet count is the result – see below left. The reason is obvious given the response curve; the second dose is supplied at the peak of the curve but the current dose will not take effect for another two weeks.
We can turn this into a closed loop system by adding feedback in the form of platelet measurements taken every two weeks which inform the dose to be administered. Unfortunately this doesn’t help at all because the measurements are always synchronised to the peak of the response curve which results in:
- An artificially high level measured resulting in a low dose administered
- Dose administration at the ‘wrong ‘time in the rhythmic cycle
- The establishment of a wildly oscillating count pattern
This problem is very well known in engineering circles and is responsible for all sorts of problems including otherwise well constructed bridges for example that become dangerously unstable when even pedestrians walk across them.
The solution though is equally well known which is simply to ensure that the feedback is out of phase with the characteristic frequency of the system. Above right we see the effect of performing the exact same procedure but on a weekly basis instead of bi-weekly. A damping effect is seen on the response curve and the desired stable target output is achieved.
Note that even daily variations in platelet count are now negligible despite doses occurring only one a week. The system is controllable with remarkably little intervention.
Complex stable systems are now easily constructed by connecting together multiple control systems in various ways.
The fact that each individual sub-system has the tendency to stabilise makes them eminently suitable as building blocks for larger systems and practically guarantees stability of the system as a whole.
Even a few simple connected components can give surprising output. Physicists will typically test the system for its response to spike inputs (single dose) and wave-form inputs (roughly equivalent to regular doses) and can easily expect the following patterns as a result:
- Spike input → Spike output
- Spike input → Spike output with delay
- Spike input → Spike followed by a smooth decay curve
- Spike input → Spike followed by oscillating decay curve
- Wave input → Wave output
- Wave input → Wave output but at a phase lag
- Wave input → Constant output
- Wave input → Wildly fluctuating output (resonance)
It is clearly not possible to deduce the whole output response from just one measurement in many of these cases. Moreover, extreme caution needs to be taken even when performing a large number of experiments as in some of the above the output is not statistically correlated to the input even though the relationship can be said to be ‘causal’.
A temperature control system is shown here as a collection of connected feedback control systems and we can imagine such a structure in the human body. Of importance is the existence of a single input (bottom left) representing the desired temperature and all the rest of the diagram will automatically adjust to that goal.
This is a neat solution then to the problem of dynamic stability whereby the body needs to maintain a constant temperature whether standing in the snow or jogging in summer heat but occasionally will need to readjust to a new temperature. In the case of a fever state, say, it will need to attain and maintain a higher yet stable temperature again in the presence of other physiological variations.
Many disease states show typical patterns of temperature variation and we can now see why – it is because the system for maintaining temperature is independent of almost all other input so the same physiological system is in effect no matter what the disease.
Blood pressure is controlled by a control system (right) which at first sight looks to be too complex to understand but upon closer examination reveals itself to be stabilised by a single metric, the signal from the baroreceptor neurons which measures blood pressure in the arteries.
This is the single point of feedback in the entire system and information from it will be used by the cardiovascular centre to adjust the whole system to an appropriate output pressure. Treatments for high blood pressure reflect this fact by, for example, acting upon the signal from these receptors to control the measurement of the pressure rather than acting upon the heart directly [MedlinePlus].
Note however that the system as a whole is neither observable nor controllable via this parameter alone. We can deduce hardly anything about the rest of the system by just measuring blood pressure and nor can we expect to control the inner state of the system as a whole. Attempts to fix the problem by blocking the baroreceptor signal are really just wall-papering over the problem and hoping for the best; the system is now trying to run on less information than it was before the intervention!
In attempting to fix the problem by altering the baroreceptor signal, an implicit assumption is made that this is the part that has gone wrong, that the rest of the system is in good working order and can somehow adapt to the incorrect information.
It may well be that the desired outcome (lower blood pressure) is achieved but the root cause has not been addressed.
Cell division is a highly organised process that uses information from multiple checkpoints to provide feedback. Proteins are responsible for monitoring the progress of the cell cycle, making decisions as to the integrity of the process and terminating it if necessary. The same mechanisms are present in almost all organisms from bacteria up to humans.
So closed loop control systems are at the very basis of physical life processes from the cellular up to the organ wide scale. Moreover, they can also be seen as forming the basis of consciousness and behaviour as without feedback no action or thought has any particular ‘meaning’ and cannot therefore be preferred over any other.
Shiva Ayyadurai has noticed parallels between elements of Ayurvedic medicine and the various components of closed loop systems and has even gone so far as to suggest a novel way of characterising personality types by the degree to which they are strong or weak on processing, feedback, robustness etc.
Dr Shiv’s Rosetta Stone of Siddha and Ayurveda:
Engineering
Input
Output
Transport
Conversion
Storage
Goal
Controller
Sensor
Disturbance
Siddha and Ayurveda
Karma
Karma-Phal
Vata
Pitta
Kapha
Sankalpa
Manas
Indriyas
Vikaras
Different sources have slightly different formulations of Ayurveda but the concept of a modular hierarchy of control systems is ever present:
“The three humours; Vata dosha, Pitta dosha and Kapha dosha are collectively called as Tridoshas and they control the basic physiological functions of the body along with five sub-doshas for each of the principal doshas.” – NIH
Dr. Shiv has also pointed out that the whole of human society actually runs on a closed loop system with behaviour being driven by ‘policy’ and financial incentive and feedback being provided by the many monitoring organisations run by world governments.
Supplementation and medication. There seems to be an overwhelming impression that the functions of pharmaceuticals are largely independent of each other, that there is such a thing as a ‘safe’ dose and that a higher dose has a more pronounced effect whilst at the same time acknowledging phenomena such as allergic reactions and tolerance to repeated doses. Confused.
The view of the body as a collection of connected intelligent systems helps to clarify the situation somewhat. All inputs to this system are interpreted by that system as a whole and the effect observed will be the response of the system as a whole. The overall character of the regulatory system in general is to stabilise in the presence of various disturbances or to put it another way, to resist medication in normal circumstances.
No part of the system is really independent of any other part and if we don’t know the precise function of a substance within the network then we cannot sensibly predict the effects and even large scale studies are just ‘black box’ science. Experimenters are treating the body as if it were a sealed box, inputting data at one end whilst making observations of outputs at the other and then trying to work out what is going on in between.
In the specific cases of vitamins D and C the supplements (D3 and ascorbic acid) which are held to contain the ‘active’ ingredients turn out to have toxic effects whereas the whole-vitamin complexes obtained naturally via sunlight or oranges have the desired effect of maintaining the required network subsystem as a whole. Vitamin D levels
Final thoughts.
Network systems are not necessarily controllable or observable from a finite subset of observations so it may well turn out that there is a component to health and disease that is simply not accessible to us.
Dose and response can be causally related even though there is no statistical correlation between them.
Their omnipresence and interconnectedness suggests that the characterisation of closed loop control Systems as the Basic Units of Biological Organisation is not too fanciful.
References:
Biological Control Systems: Systems Biology of Diseases and the Design of Effective Treatments – Babatunde A Ogunnaike
https://ceat.okstate.edu/che/site_files/docs/babatunde-a-ogunnaike.pdf
Editorial: Biological Control Systems and Disease Modelling
Authors: Babatunde A Ogunnaike, Julio R Banga, David Bogle, Robert Parker.
https://discovery.ucl.ac.uk/id/eprint/10172165/
Open loop vs. closed loop control systems (with Xcos simulations) – X-Engineer
https://x-engineer.org/open-loop-vs-closed-loop-control-systems/
The system and revolution – V A Shiva
Cascade Control – Science Direct
https://www.sciencedirect.com/topics/engineering/cascade-control
A glimpse of Ayurveda – The forgotten history and principles of Indian traditional medicine – Jaiswal, Williams
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5198827/
Dosha brain‑types: A neural model of individual
differences – Travis, Wallace
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4719489/pdf/JAIM-6-280.pdf
Baroreceptors – Maggie Armstrong; Connor C. Kerndt; Ross A. Moore.
https://www.ncbi.nlm.nih.gov/books/NBK538172/
Blood Pressure Medicines – MedlinePlus
https://medlineplus.gov/bloodpressuremedicines.html
Observability – Wikipedia
https://en.wikipedia.org/wiki/Observability
Cell Cycle Regulation by Checkpoints – Barnum, O’Connel
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4990352/