Extracting a Behavior List

Greetings, John Lighton here and I’d like to show you how we can extract a list of behaviors from a Promethion metabolic phenotyping recording. Here we’re tracking four different sensor outputs on this graph. We’re tracking body mass in black, food intake in red, running on the running wheel in gold, and water intake in blue.

So, for example, here the animal is in the body mass habitat. It leaves the habitat, takes a drink of water, and then gets on the running wheel. Then gets back in the body mass habitat, eats – you can see here the mass of the food hopper before and after the feeding event – takes a large drink of water, gets back into the body mass habitat, gets a bit restless which you can tell from the disrupted mass trace, takes a bit of a drink of water, gets back on the running wheel and so on.

The mouse alternates between different behaviors. Sometimes it does not interact with any of these sensors we’ve mentioned; just the X-Y position sensor and we refer to that as a lounge or roam. And anyway, here the mouse is going through different behavior transitions, and we can track every single one of them with one-second resolution.

So here we go transitioning from behavior to behavior. And from this information we can create a list of all the mouse’s behaviors one after the other. For example, here at this time it did 12 seconds of drinking water; took in 55 microliters. At this time, it did 205 seconds of running for 171 revolutions on the running wheel, and then it at this time did 88 seconds of lounging around in its little habitat where it weighed 25.63 grams. Additionally, you can actually take any of the other sensor’s outputs measured by the system, such as for example energy expenditure, RER and what have you, and add columns in which you can see each behavior linked to another variable.

Thanks for listening.