June 6, 2018
This post is relevant for MedPC users who also use MatLab or Python for data analysis.
We recently became aware that many MedPC users are not saving precise times for behavioral events from their experiments. A method called time-event codes was worked out around 2000 by Russ Church and his group at Brown, working with MedAssociates. Marcelo Caetano, a former postdoc in the Laubach Lab at Yale, incorporated this approach into an existing MatLab function (MedParse, written by Kumar Narayanan during his PhD training in the Laubach Lab at Yale). More recently, the code was ported to Python by Kyra Swanson, a Phd student in the Laubach Lab at American University. It is available at https://github.com/LaubachLab/MedParse. MedPC code for saving precise times of behavioral events (example in MedPC Template) and MatLab and Python functions are provided that convert MedPC data (see the template) into “time-event codes,” i.e., a two-column matrix with times (column 1) and events (column 2).
May 21, 2018
Meaghan Creed has developed a novel device for assessing preferences by mice among fluids in their homecages, i.e. two-bottle choice test. She shared the design on http://hackaday.io and contributed the summary of it below.
Often in behavioral neuroscience, we need to measure how often and how much a mouse will consume multiple liquids in their home cage. Examples include sucrose preference tasks in models of depression, or oral drug self-administration (ie. Morphine, opiates) in the context of addiction. Classically, two bottles are filled with liquids and volumes are manually recorded at a single time point. Here, we present a low-cost, two-sipper apparatus that mounts on the inside of a standard mouse cage. Interactions are detected using photointerrupters at the base of each sipper which are logged to an SD card using a standard Arduino. Sippers are constructed from 15 mL conical tubes which allows additional volumetric measurements, the rest of the holding apparatus is 3D printed, and the apparatus is constructed with parts from Arduino and Sparkfun. This automated approach allows for high temporal resolution collected over 24 hours, allowing measurements of patterns of intake in addition to volume measurements. Since we don’t need to manually weigh bottles we can do high-throughput studies, letting us run much larger cohorts.
This is designed such that each set of 2 sippers uses its own Arduino and SD card. With a bit of modification to the code one Arduino Uno can be programmed to log from 6 cages onto the same SD card. Arduino compatible boards with more GPIOs (like Arduino Mega) can log from up to 56 sippers on one Arduino.
May 15, 2018
We developed a toolbox for videographic processing of head-fixed rodent behaviors. It extracts the principal components of the animal’s behavior, either in a single movie, or across movies recorded simultaneously. Several regions of interest in the movie can be processed simultaneously (such as the whisker pad or the nose). We found that the behavioral components from the full face of the mouse predicted up to half of the explainable neural activity across the brain (see https://www.biorxiv.org/content/early/2018/04/22/306019).
In addition to extracting movement components, it can compute the pupil area of the rodent using a center-of-mass estimation. Also, in experiments in which the mouse sits on a surface with texture, the software can estimate the running speed of the mouse. The software is available here (https://github.com/carsen-stringer/FaceMap).
April 2, 2018
Check out the Ethoscopes platform!
Ethoscopes enable high-throughput analysis of behavior in Drosophila and other animals for <$100. The system is capable of real-time video tracking, is based on raspberry pi, and even has its own R package for data analysis. All software and build specifications are available at http://lab.gilest.ro/ethoscope.
March 9, 2018
O’Leary and colleagues describe an open-source touch-screen for rodent behavioral testing. The manuscript is well documented and includes all of the parts needed to build the system on your own. Very useful methods for testing cognitive function and relating findings across species (rodents, primates, humans). Congrats to the authors on setting a high standard for open-source neuroscience!
O’Leary, J.D., O’Leary, O.F., Cryan, J.F. et al. Behav Res (2018). https://doi.org/10.3758/s13428-018-1030-y
March 8, 2018
Robyn A. Grant, from Manchester Metropolitan University, has shared the following on Twitter regarding the development of the LocoWhisk arena:
“Come help me develop my new arena. Happy to hear from anyone looking to test it or help me develop it further.”
The LocoWhisk system is a new, portable behavioural set-up that incorporates both gait analysis (using a pedobarograph) and whisker movements (using high-speed video camera and infrared light source). The system has so far been successfully piloted on many rodent models, and would benefit from further validation and commercialisation opportunities.
Learn more here: https://crackit.org.uk/locowhisk-quantifying-rodent-exploration-and-locomotion-behaviours
March 1, 2018
From the Kravitz lab at the NIH comes a simple device for dispensing pre-measured quantities of food at regular intervals throughout the day. Affectionately known as “SnackClock”, this device uses a 24-hour clock movement to rotate a dispenser wheel one revolution per day. The wheel contains 12 compartments, which allows the device to dispense 12 pre-measured “snacks” at regular 2 hour intervals. The Kravitz lab has used this device to dispense high-fat diet throughout the day, rather than giving mice one big piece once per day. The device is very simple to build and use, requiring just two 3D printed parts and a ~$10 clock movement. There is no microcontroller or coding required for this device, and it runs on one AA battery for >1 year. The 3D files are supplied and can be edited to fit SnackClock in different brands of caging, or to adjust the number of snack compartments. With additional effort the clock movement could be replaced by a stepper motor to allow for dispensing at irregular or less frequent intervals.
An interesting summary of recent methods for monitoring behavior in rodents was published this week in Nature.The article mentions Lex Kravitz and his lab’s efforts on the Feeding Experimentation Device (FED) and also OpenBehavior. Check it out: https://www.nature.com/articles/d41586-018-02403-5
February 6, 2018
Brian Isett, who is now at Carnegie Mellon, has kindly shared the following tutorial regarding the creation and implementation of a Rodent Running Disk he designed while at University of California, Berkeley.
“Awake, naturalistic behavior is the gold standard for many neuroscience experiments. Increasingly, researchers using the mouse model system strive to achieve this standard while also having more control than a freely moving animal. Using head-fixation, a mouse can be positioned very precisely relative to ongoing stimuli, but often at the cost of naturalism. One way to overcome this problem is to use the natural running of the mouse to control stimulus presentation in a closed-loop “virtual navigation” environment. This combination allows for awake, naturalistic behavior, with the added control of head-fixation. A key element of this paradigm is to have a very fast way of decoding mouse locomotion.
In this tutorial, we describe using an acrylic disk mounted to an optical encoder to achieve fast locomotion decoding. Using an Arduino to decode the TTL pulses coming from the optical encoder, real-time, closed-loop stimuli can be easily presented to a head-fixed mouse. This ultimately allowed us to present tactile gratings to a mouse performing a whisker-mediated texture discrimination task as a “virtual foraging task” — tactile stimuli moved past the whiskers synchronously with mouse locomotion. But the design is equally useful for measuring mouse running position and speed in a very precise way.”
The tutorial may be found here.
Isett, B.R., Feasel, S.H., Lane, M.A., and Feldman, D.E. (2018). Slip-Based Coding of Local Shape and Texture in Mouse S1. Neuron 97, 418–433.e5.
January 8th, 2018
The de Bivort lab and FlySorter, LLC are happy to share on OpenBehavior their open-source Drosophila handling platform, called MAPLE: Modular Automated Platform for Large-Scale Experiments.
Drosophila Melanogaster has proven a valuable genetic model organism due to the species’ rapid reproduction, low-maintenance, and extensive genetic documentation. However, the tedious chore of handling and manually phenotyping remains a limitation with regards to data collection. MAPLE: a Modular Automated Platform for Large-Scale Experiments provides a solution to this limitation.
MAPLE is a Drosophila-handing robot that boasts a modular design, allowing the platform to both automate diverse phenotyping assays and aid with lab chores (e.g., collecting virgin female flies). MAPLE permits a small-part manipulator, a USB digital camera, and a fly manipulator to work simultaneously over a platform of flies. Failsafe mechanisms allow users to leave MAPLE unattended without risking damage to MAPLE or the modules.
The physical platform integrates phenotyping and animal husbandry to allow end-to-end experimental protocols. MAPLE features a large, physically-open workspace for user convenience. The sides, top, and bottom are made of clear acrylic to allow optical phenotyping at all time points other than when the end-effector carriages are above the modules. Finally, the low cost and scalability allow large-scale experiments ($3500 vs hundreds of thousands for a “fly-flipping” robot).
MAPLE’s utility and versatility were demonstrated through the execution of two tasks: collection of virgin female flies, and a large-scale longitudinal measurement of fly social networks and behavior.
Links to materials:
Raw data and analysis scripts
De Bivort Lab Site