by Andre on 13 November 2014
The OpenWorm team is trying to simulate an entire animal. It’s a big goal, but C. elegans is the right place to start. I had the pleasure of meeting a good portion of the team last week in London at their meeting and I’m definitely impressed with what they’ve managed to do so far. You can read their latest (and I believe first!) update in Frontiers in Computational Neuroscience here.
You can find out more on the page of their (extremely successful) Kickstarter campaign. The emphasis there is on something concrete that they can give back to contributors, but the longer term goal remains the development of a flexible platform for worm simulations. The most exciting aspect for me is the behavioural validation. Basically, to check whether the simulation is working and to optimize parameters, they will compare the output to known worm behaviour.
In principle, that’s not much different from comparing a mutant worm to the wild type, which I spend a lot of my time thinking about. We’ve recently sent them the data from Ev’s behavioural database paper and they’ve kindly agreed to serve the data, including the videos, which we are currently only doing for the segmented videos via our YouTube channel.
Our next generation tracker is going to generate some even better data both for our purposes in behavioural genetics, but also for OpenWorm’s simulation because we’re going to include a few stimuli rather than just looking at spontaneous behaviour. That kind of input-output data will be more useful for model optimization. Of course, what they really want is comprehensive optogenetics and imaging data coupled with quantitative behavioural data. That’s not quite available yet, but some groups are getting very close.