We are developing cyberinfrastructure tools for addressing emerging needs of big data in neuroscience based on use cases (compute/data intensive) identified through user surveys at MU. CyNeuro currently features some of the use cases for deploying cyberinfrastructure to faciliate use and evaluation of the tools in the broader community. Read more.
CyNeuro hosts exemplar workflows and data sets that we are developing for exploring the potential of cyber and software automation in neuroscience. We are connecting to and leveraging local MU resources, as well as Neuroscience Gateway resources to scale research productivity, and develop large-scale training platforms. Read more.
CyNeuro hosts training modules that are by-products from student-driven projects at MU in a course on “Cyber and Software Automation in Neuroscience”. NSF and NIH sponsored training program needs are being met through the training content development and online dissemination using Jupyter Notebooks. Read more.
Chatbot Driven Neuron Single Cell Simulation on CyNeuro Portal
Guided user interface for simple single cell neuron simulation. The user will be navigated through a series of steps of simulation by a chatbot. At each step, the chatbot guides the user by providing useful information through a dialog relating to the simulation.
The guided user interfaces for the Neuron Simulations were driven by a set of responsive/dynamic forms and a context-aware chatbot. This chatbot is context-aware i.e., it responds to the user questions about the workflow based on the profiled user quadrant. Users are able to navigate through the user interface screens of a neuroscience workflow, while simultaneously interacting with the chatbot. The chatbot is equipped to help the user enter the parameter values that are needed to run their workflow, as necessary.
Students, teachers and faculty members with a varying level of expertise can interact with our bot to get help with neuron simulations and experiments. As the user interacts with the bot, we collect basic and advanced information such as e.g., end goal of the user, how much time can he spend to learn, his/her background information to build a user profile. The user profile is beneficial to generate contextually correct responses to the users based on their proficiency level in cyberinfrastructure and neuroscience.
Neuron Simulation using Jupyter Notebook
Jupyter Notebooks are documents that contain computer code and rich-text elements (e.g., equations, figures, web links) in order to enable the user (i.e., an instructor or student) to inspect the state of an execution in real-time. This server-client application has the flexibility to edit and execute a custom notebook on a local desktop without Internet access or via web-access on a remote server. These features allow the simulations to be modified and communicated in an understandable and reproducible manner and can be used to remotely deliver instructional material to a large group of remote students.
We developed Jupyter Notebook instances for exemplary teaching and research use cases of neuroscience users. Users can use pre-built notebooks to learn how to model a neuron cell using NEURON python tool. By obfuscating the underlying Python code, the NEURON-based notebooks allows the non-programmers to focus on experimenting with the parameters for neuron simulations. This approach also enables the instructors to focus on the learning material and teaching objectives during the class, instead of addressing trivial issues such as environment configuration, credentials for software access, or other installation problems.
Note: Please email Dr. Satish Nair for access to Jupyter Notebook server.
A simple web-based user interface for submitting jobs to NSG. Users can submit any modeling and simulation jobs to NSG from CyNeuro portal using tools available on NSG. Researchers who want to access and run computational software applications on high-performance computing resources on NSG portal can use this as an early testbed to get some familiarity with NSG tools before actually scaling their simulations.
NSG-R Jupyter Notebook
Pre-configured Jupyter notebooks for submitting neuron jobs to <a href=”https://www.nsgportal.org/” target=”_blank “>NSG</a>. Users can use this notebook to submit modeling and simulation jobs. Researchers who are familiar with neuron job submission can use these notebooks to access and run computational neuroscience software applications on high-performance computing resources. These notebooks provide a convenient way to run neuroscience applications programmatically on large HPC resources.
Acknowledgement: This notebook is adapted from iPython tutorial notebook on NSG portal. To know more about this click here.
Zebrafish Usecase using Jupyter Notebook
ModelDB provides an accessible location for storing and efficiently retrieving computational neuroscience models.
NeuroMorpho is a centrally curated inventory of digitally reconstructed neurons associated with peer-reviewed publications. It contains contributions from over 300 laboratories worldwide and is continuously updated as new morphological reconstructions are collected, published, and shared.
Soybean Knowledge Base (SoyKB) is a comprehensive all-inclusive web resource for soybean translational genomics. SoyKB is designed to handle the management and integration of soybean genomics, transcriptomics, proteomics and metabolomics data along with annotation of gene function and biological pathway.
KBCommons, is a platform that automates the process of establishing the database and making the suite of informatics tools and visualization capability easily available for other organisms via a dedicated web resource.
Please email Dr. Satish Nair for access.
‘SimAgent’ directly interfaces the student laptop with NSG resources to run network simulations. SimAgent has two core functions, automated job submission and parameter sweep. The automated job submission feature is a point and click interface that accepts any neuron or python program directory, submits the program to run remotely and watches it until completion with live updates to the user. The parameter sweep feature allows the same functionality with the added ability to specify sections of code to automatically change with each run. Users can specify a range of values for a parameter to take on, run each simulation in a parallel configuration and determine the optimal output for their needs. It currently supports connections to the NSG-R restful API and connections using SSH to servers running Slurm.
SimBuilder is an easy to use desktop application that allows users to build large scale networks in Neuron. Utilizing Marianne J Bezaire’s CA1 model (https://github.com/mbezaire/ca1) as a backend, SimBuilder generalizes the process allowing users to specify their own network requirements.