Project reference: 1505
A clear understanding of the spatio-temporal dynamics of olfactory bulb mitral cells in representing an odor input is very difficult to achieve experimentally; for this reason, the activity-dependent effects of olfactory bulb network self-organization generated by the mitral-granule cell microcircuits remain poorly understood. To deal with this problem, we have constructed olfactory bulb microcircuits using realistic three dimensional (3D) inputs, cell morphologies, and network connectivity.
The computational approach is based on a novel NEURON + Python fully integrated parallel simulation environment, that can be applied to other regions as well. A population of approximately 700 synthetic mitral cells and 120,000 granule cells were randomly connected using a collision detection algorithm. Intrinsic optical imaging data obtained during presentation of 19 natural odors was used to drive self-organization of the network under different conditions of odor inputs.
The results provided new insights into the relations between the functional properties of individual cells and the network in which they are embedded, making experimentally testable predictions on odor processing and learning.
Project mentor: Michele Migliore
Site Co-ordinator: Francesca Delli Ponti
Manage parallel applications for 3D visualization. Increase student’s skills on parallel computing. Manage neurological data.
Student Prerequisites (compulsory)
MPI, Blender, Paraview, Python, C/Fortan
Student Prerequisites (desirable)
NEURON, RTNeuron, POV-Ray
- Week 1: Introduction to CINECA systems, and small tutorials on parallel visualization
- Week 2: Assess modifications needed for the reading in and interpretation of datasets and the additional functionality required to output movies
- Week 3: Deliver Plan at the end of weekWeek 4, 5, 6: Production phase
- Week 7: Preparation of the final movie
- Week 8: Preparation of the final movie. Write the final Report.
Final Product Description
The purpose is to set up a tool or scripts for visualizing data coming from scientific program with all the problem of converting data, insert camera and light and render a realistic scene. All these work will be described on a final video and a small paper will be produced.
Adapting the Project – Increasing the Difficulty
An interface user friendly for using the realized scripts also integrated to existing software if needed. Furthermore, a 3D model could be integrated in a web page with real time navigation.
Adapting the Project – Decreasing the Difficulty
The project could be stopped on scripts that convert data to input format file of Blender and Paraview or other useful software for better be visualized.
The software that we will use are open source: Blender and Paraview. Both are installed on our supercomputer and the students need only the account to access to them.