Last week I went along to a seminar given by Mike Vella at the Behavioural and Clinical Neuroscience Institute here in Cambridge. Mike spoke about the Open Worm project which aims to simulate a whole C. Elegans organism in silico. It’s also a great example of Open Science.
C. Elegans is a nematode worm, about 1mm long and transparent, which lives in the soil and feeds on decaying vegetable matter. It is probably the single most well documented organism in biology. Web sites such as WormBase and WormAtlas are brimful of information from its genome to the position and function of every one of its roughly 1000 cells. Just over 300 of those cells are neurons, which is enough for C. Elegans to exhibit some interesting behaviour: it can recognise and move towards food, it will detect and move away from chemical toxins, it can learn (some very simple) behaviours, it even ‘sleeps’. C. Elegans is the only species for which a full connectome has been published: all the neuronal connections – neuron to neuron and neuron to other cell types such as muscle – have been mapped. The full cell lineage is also known, that is how the 1000 cells in an adult are related to and derived from the original egg through the development of the embryo worm. This abundance of data suggests that if we want to model a whole organism computationally, C. Elegans is a great place to start.
The Open Worm project aims to do just that. It is a loose and widely distributed collaboration involving computational neuroscientists, professional software engineers and biologists in the US, UK, Italy & Russia. They’re keen to recruit new project members. It’s not currently funded by any research grant but is run by the participants in their “spare” time. All source code, data and publications are Openly available. They make great use of online tools such as Google Plus, GitHub and Amazon Web Services.
So far they’ve built a 3D visualisation using WebGL – the Open Worm browser – which shows the various structures in the worm down to the cellular level. They also have a mechanical model which simulates the physical motion of the worm. The simultaneous firing of the 302 neurons constituting the “brain” of C. Elegans has been simulated. The next step is to model the connection of a neuron to a muscle cell.
Why bother constructing a worm in silico? After all, it’s pretty easy to work with real life C. Elegans in the lab. Well, one of the reasons for constructing a computer model is that it helps organise what we know about the subject. There’s still an awful lot we don’t know about C. Elegans and attempting to simulate it helps identify those gaps. Even more ambitiously, it’s a tentative first step towards more complex simulations, eventually including the human brain.
People have tried and failed in the past to simulate C. Elegans. The Open Worm project may or may not succeed though I think that adopting an Open approach tips the balance in their favour and has given them an important edge that previous work has lacked.