Biomimetics and The Origin of Nervous Systems: Nature’s Patents - Part II
We ended the first instalment of this series with an outlook on the fusion of biomimetics and comparative genomics. In July, I presented a paper at the “Living Machines 2017” conference in Stanford that proposed a new method to achieve just that — living machines. The published paper was titled: “Introducing Biomimomics: Combining Biomimetics and Comparative Genomics for Constraining Organismal and Technological Complexity”.  The premise of this newly developed method is quite simple: with the use of bioinformatics to compare sequenced genomes from an array of species, we can learn to read ‘Nature’s Patents’. This will allow us to extract a wealth of ‘design blueprints’ that have the potential to inform, and essentially transform, technological development across a range of disciplines.
However, this task is most definitely not as straightforward as it may sound. Finding solutions to biomimetic problems requires the concentrated efforts of computational biologists, bioscience researchers from different academic backgrounds, industrial and product designers, as well as a host of engineering and scientific experts with domain knowledge in their respective fields.
We envision that the applications of this work will be extensive and far-reaching — from materials science and biomedicine to architecture, deep-sea exploration and possibly entirely new computing architectures. In essence, we are convinced that it is possible to ‘biomimetize’ our technology, and by extension modern life, in unforeseen ways.
Analogue vs. Digital Neuromorphic Computing of the Future
In the field of biomimetics, a growing number of researchers are trying to establish computer-aided methods for optimizing design processes and facilitating breakthrough discoveries in bio-inspired engineering and technology. Another goal of such efforts is to promote collaborative design thinking among engineers, biologists and other experts. Biologists are turning into practical-minded engineers, while engineers are starting to familiarize themselves with the way Mother Nature “thinks” and organically creates ingenious solutions to problems that man-made technology would struggle to crack.
“Amazingly, modern sequencing technology is not only able to decipher the DNA code. The intermediate mRNA molecules can be read as well, thereby providing important insights on the transcriptional state of a cell.”
While biomimetics has already been defined in part I, it is still important to provide a succinct understanding of what is meant by “comparative genomics”. Essentially, genomics is concerned with deciphering the totality of a given species’ genetic code through the use of high-throughput sequencing technology. The generated raw sequence data is then put together by automatized bioinformatics pipelines and translated to its corresponding amino acid (i.e. protein) sequences which are easier to work with.
It is important to note, that the entirety of a given organism’s DNA is not always translated into protein; a considerable portion is needed for regulatory purposes and these genomic areas can be mapped as well. Fascinatingly, modern sequencing technology is able to not only decipher the DNA code (and protein sequence) but also to read the intermediate mRNA (messenger RNA) molecules, thereby providing important insights on the transcriptional state of a cell. In effect, we can get a snapshot of exactly what type of proteins a cell produces at a specific moment in time. Genomics is therefore complemented by transcriptomics.
Once computational biologists have put together the genome under investigation, the interesting part begins. So-called ‘sequence alignments’ are used to map genomic regions of interest from one organism to another (or a multitude of them). This allows them to compare the degree to which a particular region has been conserved or has diverged over evolutionary time (Figure 1).
Even if a gene has greatly diverged, in most cases it is still recognizable across vast evolutionary distances, for example between jellyfish and Homo sapiens. The same genes in different species are called homologous. Other, more drastic, ways in which genomes evolve, include the duplication of a gene or whole-genome, gene losses, protein domain shuffling, gene birth and other forms of rearrangement that occur on the genome level.
Figure 1: This chart illustrates the different scales of comparison that can be covered by modern comparative genomic approaches. On every level, we can discover things that organisms have in common and features that are specific to a particular evolutionary lineage. Credit: NIH
Nervous Systems as Multiples
The central nervous system (CNS), embodied by the intricate brains of mammals, is generally regarded by scientists as the most complex form of biological organisation. Although not all animals have brains or “centralized” neural organisation per se, the great majority of animal species do have nervous systems in one form or another.
“The great question of evolutionary neurobiology now is whether a primitive nervous system originated only once or emerged independently within different branches of the tree of life.”
The great question of evolutionary neurobiology now is whether a primitive nervous system originated only once and, subsequently, complexified during the ca. 600 million year history of animal life — or, alternatively, whether nervous systems emerged independently within different branches of the tree of life. In the latter case, neural computation represents a multiple — a repeated natural solution to a complex problem (as defined in part I of this series). Multiples by their very nature entail multiple solutions to a given problem by means of functional equivalence.
In the broadest sense, functional equivalence means that a given function (top-down level) can be instantiated by a variety of lower-level agents. Bioluminescence is a fitting example of functional equivalence. In nature, the higher-level goal of producing biological light is exemplified by a plethora of molecular systems that are capable of carrying out this specific function. These systems originated independently from one another via convergent evolution.
On an abstract systems theoretical level, the convergent evolution of organismal complexity is the result of biomolecular networks that are organized and structured by information hierarchies via top-down causation. The emergence of fundamental network properties, such as modularity and the functional equivalence classes of lower-level operations or subroutines (which occur in both biological and technological systems), can be explained as a corollary of top-down organized information hierarchies.
Top-down causation refers to the causal role of information in living systems. More specifically, it describes the process by which higher levels of organisation in structural hierarchies constrain the dynamics of lower levels of organisation. Top-down theorists propose that the organisation structure and function of living systems cannot be completely explained by a reductionist “bottom-up” approach. These thinkers include Sara I. Walker, a physicist at Arizona State University who is leading their “Emergence” group and has contributed a number of highly innovative and inspiring papers.
Figure 2: Information hierarchies in biological systems. Information can flow in both directions: from the molecular level upwards to higher-order cellular structures (ie. cells, tissues, organs). These, in turn, constrain and influence dynamics at lower levels via top-down causation and information control. Subordinate only to the organism as a whole, the nervous system is the central top-down causative agent in higher animals.
In most reductionist approaches to biology and science, it is assumed that purely physical effects determine the dynamics of lower levels of organisation and, by extension, strictly govern interactions that occur at higher levels as well. However, information can acquire a causally efficacious role in physical systems without violating the principle of physical “closure”. In fact, an emerging school of thought in evolutionary biology is advancing the hypothesis that the transition from non-life to life, abiogenesis, can be understood as a transition in causation and information flow.
Informational takeover events appear to have occurred at all major evolutionary transitions (and technological for that matter — the emergence of the Internet being the most obvious example). At any given higher level of biological complexity, there are novel top-down informational hierarchies which define lower-level molecular, cellular or organismal operations (Figure 2). This idea is closely related to the concept of functional equivalence classes:
“Top-down causation operates through functional equivalence classes. Functional equivalence occurs when a given “higher-level” state leads to the same high-level outcome, independent of which “lower-level” state instantiates it. Equivalence classes are defined in terms of their function, not their particular physical instantiation: operations are considered (functionally) equivalent (i.e., in the same equivalence class) if they produce the same outcome for different lower-level mechanisms.” 
A better understanding of the functional equivalence classes that make up varying types of neural organisations and computations — specifically, in terms of molecular physiology and systems architecture across animal complexity — could facilitate the design of robotic and computational platforms that truly fulfil their “natural” purpose.
Nature-Inspired Robotic and Computational Platforms
A striking and very much applicable example here could be a long-distance soft-robotic jellyfish-like AUV (autonomous underwater vehicle) equipped with a new kind of hydrogel-based computational apparatus. The hardware requirements for such an implementation would be completely different from the type of CMOS (Complementary metal–oxide–semiconductor) chips used in mobile or desktop applications. Remarkably, Nature has already devised designs for the different environments in which diverse ways of computing information are needed.
Clearly, there are fundamental similarities in the way nervous systems work across animal complexity as well. However, comparative genomics allows us access to the neural black box which reveals how the same function (neural computation) is carried out by different molecular systems in an increasing number of model and non-model organisms. All evidence points toward the fact that nervous systems have multiple origins and, as such, present one of the most striking examples of convergent evolution.
It is quite extraordinary to consider that the most complex known information processing system in the Universe has multiple origins. Nervous systems are responsible for the top-down control of the organism’s physiological body functions, energy metabolism, movement, integration of different sensory modalities and, finally, its advanced cognitive capabilities (ie. perception, memory and thinking).
The holy grail of AI research and advanced R&D in robotics is to reverse engineer this natural system and essentially design a machine that is capable of performing brain-like functions. Although, “brain-like” functions, in this case, refers to a nervous system function that is more similar to that of an insect or fish, rather than human-like. Swarm intelligence is a key example of where a better understanding of the neural basis of such complex behaviour, in an organism-specific way, is necessary. Biomimomics offers exactly this kind of analysis and, furthermore, provides a workflow that greatly accelerates R&D and design processes by rendering them more efficient.
As demonstrated, nervous systems may be more akin to analogue computers of the past than current digital technology after all. Apparently, contrary to current trends and instincts within the technology, Nature is telling us to go analogue. In part III of our biomimetics series, we will delve deeper into the multiple origins of oceanic nervous systems, and discuss the latest genomic analyses that are unlocking the novel functional equivalence classes of natural computation.
 Flores Martinez C.L. (2017) “Introducing Biomimomics: Combining Biomimetics and Comparative Genomics for Constraining Organismal and Technological Complexity”. In: Mangan M., Cutkosky M., Mura A., Verschure P., Prescott T., Lepora N. (eds) Biomimetic and Biohybrid Systems. Living Machines 2017. Lecture Notes in Computer Science, vol 10384. Springer, Cham.
 Walker, I.S., “Top-Down Causation and the Rise of Information in the Emergence of Life.” Information, 2014. 5(3).
No longer relegated to the world of comic books, the amazing bionic man or woman is less sci-fi fantasy and more Barb next door. While the term ‘bionic bodies’ may conjure up dystopic images of cyborg armies amongst the paranoid and conspiratorists, innovations within biotech are revolutionising medicine and the field of prosthetics. Incredible advancements in the development of synthetic body parts has meant that the reconstruction of the human body is simpler, cheaper and, to be candid, pretty frickin’ cool. From exoskeletons and prosthetic limbs to artificial organs and deep brain implants, which can regulate abnormal brain activity caused by Parkinson and epilepsy, it is very much a case of ‘if it can be imagined, it can be created’.
Illustrations by Ilya Martynov
To contact the editor responsible for this story:
Margarita Khartanovich at email@example.com
- Joshua Lind: Best Results Will Come from Academia, Industry and Community Working Together
- Draper University: Can Innovation and Entrepreneurship Be Taught?
- How Much Can Lightning Help Upgrade a Battling Blockchain?
- Blockchain at Berkeley: From a Community Initiative to Research Projects and University Courses
- IBM Blockchain Labs’ David Gorman Talks Hyperledger and The Future of Blockchain