The thriving field of system neuroscience has enhanced the status of behavioral experiments in daily research. As the direct readout of certain brain's activities, the behavior of animals has been put at the center of the stage for 100+ years. While generations of scientists have established numerous behavioral paradigms to address the functions of diverse brain circuits, but scarce packages were developed to help people process these behavioral data, which make system neuroscience research hard for people who is not good at programming to join in.
Of course there're several excellent Python sofewares developed to address this issue, e.g, DeepLabCut. DeepLabCut provides a strong function in processing animals' movements collected by cameras (e.g., the moving trajectory of the limbs of rats) and is thus hotly welcomed by researchers. However, moving trajectory is only one of the attributes of animals' behavior, which means there are other things we could do with the behavioral data. So, we'd like to emphasize here that we are certainly not pursuing substituting these preexisting, well-developed, and powerful tools with our MazePy.
In the research field of the spatial navigation system of the brain, one common thing researchers like doing is manipulating the shape of an environment and training mice or rats to navigate or randomly forage within this environment. Tens of, if not hundreds of, environments or mazes (T maze, Y maze, 8-arm maze, linear track, L maze, 6-arm maze, pin maze, complex maze, 3D maze) have been designed in the past 50+ years to investigate the neural representation of the space.
However, packages trying and setting up to unify the analysis of different shapes of the environment are scanty. So, one of our aims is to develop a package that could well fit the analysis of most of the environments, so that researchers would no longer need to write new code to process data collected in an entirely different environment.
MazePy provides a designer GUI to assist you in designing a new environment, and functions to help you analyze the data collected in the environment:
After tens of years of rapid development, several tools have been developed to record neural activities, which could be divided, in general, into three groups:
- Genetically encoded calcium indicator (e.g., GCaMP family etc.)
- Electrodes (e.g. tetrode array, Neuropixel, etc.)
- Genetically encoded voltage indicator (GEVI)
Several useful tools have been developed to process raw neural data, e.g., suite2p, CNMF-E and numerous spike sorting algorithms. However, the one that makes people thrash out is that, generally, different analysis packages save the processing results in divergent data forms, which makes code incompatible with being used in dealing with other data forms. Here, we are trying to establish very basic but standardized data structures, to make the analysis of different forms of data more easily.
The 1.0 framework of the MazePy:
We are certainly very open to new members! Welcome every talented and skillful person to join us! Here're some basic requirements for joining our group:
- Have, at least, an 1-year experience of 'Python' programming.
- Have interests in neuroscience and enthusiasm in working with others.
If you are interested in our project, feel free to reach out to Shuyang: [email protected] (suggested) or [email protected]