Background information

Written by Super User on 27 February 2013

As Gabbiani and Cox put it, "from the simplest nervous system to that of a human, the basic principles underlying information processing appear to be universal, in spite of obvious differences in implementation" (Gabbiani & Cox, 2010). Surprisingly, even simple biological neural networks can outperform today's fastest computational systems in tasks such as sensory pattern recognition and locomotion control. Nervous systems are complex, highly parallel information processing architectures made of seemingly imperfect and slow, yet highly adaptive and power-efficient components to carry out sophisticated information processing functions. However, despite the rapidly growing body of knowledge on almost every aspect of neural function, currently no computational model or hardware emulation exists that is able to describe or even reproduce the complete behavioural repertoire of the nematode Caenorhabditis elegans, an organism with one of the simplest known nervous systems.

This is particularly surprising because C. elegans, a soil-dwelling worm with a life span of a few days, 1 mm long and 80 µm in diameter, is one of the five best characterized organisms. It is multicellular and develops from a fertilized egg to an adult worm just as a human being does. Despite its small genome (~ 10 M base pairs), there is about 40% homology to the human genome (3.2 G base pairs). The adult hermaphrodite is comprised of exactly 959 cells, including 95 muscle cells and 302 neurons. The morphology, arrangement and connectivity of each cell including neurons have been completely described1 and are found to be almost invariant across different individuals. There are approximately 7000 chemical synaptic connections, 2000 of which occur at neuromuscular junctions, and approximately 600 gap junctions (White et al., 1986)1. All of this data including the connectome, the detailed interconnectivity map of the 302 neurons through synapses, is publicly available through the Worm Atlas (Achacoso & Yamamoto, 1992; Oshio et al., 2003; Varshney et al., 2011)2. Despite its simplicity, the nervous system of C. elegans does not only sustain vital body function, but generates a rich variety of behavioural patterns in response to internal and external stimuli3. These include associative and several forms of nonassociative learning that persist over several hours (Hobert, 2003). Interestingly, many processes of learning and memory in C. elegans are highly conserved across evolution, which demonstrates that there are universal mechanisms underlying learning and memory throughout the animal kingdom (Lin & Rankin, 2010). With all of this data, information and modern computer technology at hand, it is surprising that there is yet no comprehensive artificial C. elegans emulation system from which the principles of neural information processing underlying behaviour can be derived. The Si elegans project aims to fill this gap.

Research into neural networks (NN) and artificial or computational intelligence (AI/CI) has tried for more than 30 years and not succeeded in understanding the signal processing in a simple network made of only 302 neurons. The main reason is: nervous systems work in different ways than current computers and any simulation running on them. In nature, information is processed in a highly parallel and "plastic" fashion, which cannot be sufficiently reproduced or simulated in hard-wired, serial multi-tasking or even parallel-linked von Neumann architectures.

Si elegans will innovate through several pathfinding concepts by overcoming the key limitations inherent to current computational (serial) nervous system modelling approaches. The innovation will occur by:

  1. lifting interconnection constraints in parallel computation through 3D interconnectivity concepts
  2. providing a holistic and arbitrarily configurable platform for testing neuroscientific hypotheses
  3. being completely reverse-engineerable for analysing all aspects of nervous system function
  4. being accessible (and evolvable) by any researcher through a web portal
  5. representing a paradigmatic prototype of a new generation neuromimetic computational architectures

Si elegans will not only contribute to the progress of neuroscience but may be disruptive by setting the foundation for a new era of brain-mimetic ICT. The concept of designing a truly parallel computational hardware architectures based on generic and dynamically reconfigurable neuromimetic modules that can be flexibly interconnected to result in any desired neural network circuit will serve as a basis and inspiration for creating radically new computational archetypes.

Si elegans will be provided to the scientific community through an open-access web portal to not only let anyone test their neural models and hypotheses in behavioural studies, but to actively shape the development and refinement of the computational platform. This peer-contribution concept follows the spirit of an ever growing number of successful open-access and open-contribution initiatives which have clearly demonstrated that impact is potentiated by including the scientific community and the public at large.

Because Si elegans is based on the organism C .elegans, which is one of the primary workhorses in biology and neuroscientific studies, the project will link to many existing European and international initiatives and activities.

1 The publicly available connectome is currently covering 6393 chemical synapses, 890 electrical junctions, and 1410 neuromuscular junctions.
2 All of these data have been published and are freely available, e.g., through www.wormatlas.org. A complete digital representation of the anatomy of C. elegans can be found @ caltech.wormbase.org/virtualworm. Also see: (Mailler et al., 2010)
3 Eight functional circuits have been identified, namely, (a) touch sensitivity, (b) egg laying, (c) thermotaxis, (d) chemosensory, (e) defecation, and three types of locomotion: when (f) satiated (feeding), (g) hungry (exploration) and (h) during escape behaviour (tap withdrawal) (Chatterjee & Sinha, 2007).