Nematode Nervous System: A 1-in-40-Million Design

This nematode's nervous system is perfectly wired for minimum use of materials. (click for credit)

I have been doing an “interstate book club” with one of the most brilliant people I know. She and I read the same book and call each other on a regular basis to discuss it. We are currently covering Jerry Fodor and Massimo Piatelli-Palmarini’s book, What Darwin Got Wrong. I suspect that I will do a complete review of the book at some point, but I ran across something that I found so amazing, I had to write about it today. It has to do with the roundworm known as Caenorhabditis elegans, which is pictured above. This tiny (about 1 millimeter long), transparent worm has been studied extensively. In fact, it was the first multicellular organism to have its genome fully sequenced.1

Before that happened, however, Christopher Cherniak did a detailed analysis of the creature’s nervous system. Approximately one-third of the cells in the roundworm’s body are nerve cells, so the nervous system is obviously important to this tiny animal. The system is made of clumps of nerve cells (called ganglia) in the head, tail, and scattered throughout the main nerve cord, which runs along the bottom of the worm’s body. While this system is “simple” compared to the kind of nervous systems you find in many other animals, it has served as a model for helping scientists understand how nervous systems develop and function in general.

Of course, since the nervous system has to process sensory information and control various muscle movements, the ganglia must be connected to one another, to the receptors that sense the outside world, and to the muscles that the nervous system controls. Obviously, then, there is a lot of “wiring” involved. Cherniak wanted to know what determined how this wiring was done in the animal, so he computed all the possible ways that the worm’s nervous system could be wired, given its structure and the number of components it had. His computation indicated that there were 39,916,800 ways the wiring could have been done.

Now that’s a lot of possibilities, but even back in 1994, computers could easily analyze all of them, so he used 11 microcomputers to analyze all 39,916,800 ways the nervous system could be wired. It took them a total of 50 hours to churn through the analysis, but what they found was incredible!

In comparing the different means by which the nervous system could be wired, they found that there was a big difference in the amount of wiring used. Some layouts were very efficient in how they used the wiring, and others were very inefficient. How did the real nervous system layout compare? Here is what Cherniak wrote:2

The actual ganglion order in fact appears to be the unique optimal one out of these millions for minimizing total interconnecting fiber length.

In other words, out of the nearly 40 million possibilities, the roundworm just happens to have the layout that requires the least amount of material!

Why is this important? Because it argues strongly against the neo-Darwinian view that random mutation acted on by natural selection is what determines such things. What’s the chance that random mutation would produce the most optimum design out of a possible 40 million? After all, it took 11 microcomputers (using 1994 technology) a total of 50 hours to analyze the layouts and determine the most optimum design, and they were programmed specifically for that purpose. Is it really possible that through the history of roundworm evolution, random mutation could “offer up” to natural selection the unique, best solution?

Of course, if this were the only example of such optimization, perhaps you could believe that the roundworm just got lucky. Random mutation did just happen on the most optimum layout for its nervous system. However, Fodor and Piatelli-Palmarini spend 12 pages cataloging many more examples of optimum design in living organisms. Now, of course, they don’t like to use the term “design,” because they are both atheists and don’t want to refer to any kind of Designer in their analysis. Instead, they want to believe that there are nonrandom forces at work that limit the ways in which organisms can evolve.

To me, however, the reason for the nematode’s optimal nervous system is clear: it was designed by God. As a result, you would expect the design to be elegant and efficient. Thus, the fact that it is optimized to use the least amount of material is not surprising. The same could be said of the other examples covered in the 12 pages I mentioned above. In the end, the creationist view expects a myriad of optimal designs in nature, and that’s precisely what you find. Now, of course, there are some designs that don’t appear to be optimal, but they are few and far between. In addition, some designs that initially appear suboptimal end up being pronounced optimal based on additional research. The over-arching principle in nature seems to be that of optimal design. While evolutionists might be able to eventually find some explanation around this fact, Fodor and Massimo Piatelli-Palmarini clearly show that they haven’t been able to do so yet.

In the creationist view, there is no need to explain around things like the layout of the nematode’s nervous system. It is exactly what you would expect from a creation made by the Almighty.


1. The C. elegans Sequencing Consortium, Consortium, “Genome sequence of the nematode C. elegans: a platform for investigating biology,” Science 282: 2012–2018, 1998.
Return to Text

2. Christopher Cherniak, “Component Placement Optimization in the Brain,” The Journal of Neuroscience 14(4):2418-2427, 1994.
Return to Text


  1. James says:

    This reminds me of the classic problem:

    “what is the shortest route that someone can visit each city on a list?”

    The answer to this problem is easy to verify, but from what I’ve read it is impossible to solve this problem efficiently. for example, if you have 100 cities you run up to 1.27×10^30 different distinct combinations (if i got my math right). This would take forever using trial and error, yet, it is the only way we really how to solve the problem.

    How could evolution possibly “know” what was the perfect combination?

    Also, why wouldn’t natural selection be satisfied by a combination that was close to the perfect fit? As you get closer and closer to the perfect value i assume the amount of “wiring” difference gets very small and insignificant.

    (see the ending)

    1. jlwile says:

      That’s a great point, James. In fact, it seems that bumblebees can solve simple versions of that problem.

      The authors of What Darwin Got Wrong ask the same question you asked: “Why wouldn’t natural selection be satisfied by a combination that was close to the perfect fit?” They make the strong case that natural selection should not select for the optimal solution, as the extra energy expended in the search is not worth the slight increase in survivability that results when you move from “almost perfect” to “perfect.”

  2. James says:

    Amazing. Thanks for sharing that, it shows how special we as humans are that God would create all this splendid natural beauty just for us!

  3. josiah says:

    I don’t really trust that number. There may be 40 million possible ways to link those components, but most of them are so obviously outrageous, perhaps linking sensors in the tail to the head, that they’re clearly not options.

    Moreover, from a computer science perspective this doesn’t seem all that impressive.
    The question that James refers to is called the travelling salesman problem, and is indeed a very tricky one with no known algorithms that are faster than 2^n.
    The problem here, however, is more akin to finding a minimum spanning tree. There are more efficient algorithms for solving that one.

    Moreover a computer would be trying to find the best way to link a set of pre-defined nodes one after the other. By contrast the evolution of a biological organism would be developing and placing the nodes at the same time as it selects the link between them. In some respects it’s also working “in parallel” rather than checking one edge after the other. That makes the problem completely different from the one the computers addressed.

    1. jlwile says:

      Those are excellent points, Josiah. I agree that many of the possible combinations are outrageous. Thus, they can be dismissed out of hand. However, the authors’ point is that there are all sorts of solutions that are very close to the optimal one. From a point of view of natural selection, any one of those solutions would be “good enough” for survival. Why is it that the optimal configuration is the one we see?

      From an evolutionary point of view, the nodes and the wiring are being developed together. This makes it different from the problem the computers were trying to solve. However, I’m not sure it makes it easier to find the optimal solution. Consider, for example, a “protonematode” whose nervous system is less advanced. What would that nervous system look like? It seems to me that it would have a ventral nerve cord, but not as many ganglia. Now random mutation has to produce ganglia along that cord. I don’t see random mutation being very systematic about it, so I assume one ganglion would appear in one place, and the next one would appear a random distance from the first. In this scenario, it would seem to me that a lot of extraneous wiring would have to be developed. I would think that would make it harder to find the optimum solution.

  4. BWW says:

    Very interesting. This raised a question for me: Is the “wiring” of the nervous system nearly the same for all members of C. elegans? I don’t know much about embryology or gene regulation, but I do know that some animals develop by an almost unvarying set of divisions, specializations, outpocketings, splittings, etc. How relevant is the study to all members of the species?

    1. jlwile says:

      BWW, if we ignore mutants, the wiring is identical among all members of the species. So this applies to all non-mutant members of the species.