Fruit Fly Brain Hackathon 2017


FFBH 2017

March 12, 2017

Center for Neural Engineering and Computation

Columbia University, New York, NY 10027


Overview

The goal of the hackathon is to bring together researchers interested in developing executable models of the fruit fly brain. Towards that end we will engage systems and computational neuroscientists in modeling, design, implementation and biological validation of an open-source emulation platform of the whole fruit fly brain. All hackathon participants will be provided with an Amazon Machine Image of the recently developed open-source Neurokernel platform [1] for executable fruit fly brain circuits.

The hackathon is aimed at three main groups of participants: biologists, modelers and software engineers. For biologists, the hackathon focuses on the intuitive modeling and representation of biological data, such as anatomical and recordings data of the fruit fly brain, in the NeuroArch database. For modelers, the hackathon aims at creating/modifying models of neuropils that are compliant with the Neurokernel API. For software engineers, the hackathon focuses on improving the Neurokernel platform and its API, and developing new, proof-of-concept features that are needed by biologists and modelers alike. All hackathon participants will be strongly encouraged to collaborate towards the realization of executable fruit fly brain models.

The Fruit Fly Brain Hackathon is organized in conjunction with the Columbia Workshop on Brain Circuit, Memory and Computation on March 13-14, 2017. Participants of the hackathon are welcome to attend the workshop.

[1] Lev E. Givon and Aurel A. Lazar, Neurokernel: An Open Source Platform for Emulating the Fruit Fly Brain, PLOS ONE 11(1): e0146581. doi:10.1371/ journal.pone.0146581, January 2016.

Organizers

Paul Richmond, Department of Computer Science, University of Sheffield
Adam Tomkins, Department of Automatic Control and Systems Engineering, University of Sheffield
Nikul Ukani, Department of Electrical Engineering, Columbia University
Chung-Heng Yeh, Department of Electrical Engineering, Columbia University
Yiyin Zhou, Department of Electrical Engineering, Columbia University

Registration

TBA

Lodging and Directions to Venue

TBA

Schedule

Thursday, March 12th, 2017

Detailed schedule to be announced.

Computing Resources

The hackathon relies on the Amazon Elastic Compute Cloud (Amazon EC2) as the main computing resource. Prior to attending the hackathon, participants are requested to set up an account with Amazon EC2 and to familiarize themselves with creating Linux GPU instances. Please be aware that Amazon charges for the EC2 service (pricing info can be found here). Students can obtain free credits from AWS Educate. During the hackathon high speed wired and wireless network Internet access will be provided to all participants.

To help participants get started with Neurokernel, we provide an Amazon Machine Image (AMI) with a preloaded Neurokernel. For more information about how to use this AMI and how to get started with Amazon EC2, please see this notebook.

In addition to using the Amazon EC2 service for high performance GPU computing, a limited number of Nvidia Jetson TK1 embedded development kits will be provided to participants interested in developing fruit fly brain models using embedded systems. The Jetson TK1 platform features a Tegra K1 SOC that consists of a NVIDIA Kepler GPU with 192 CUDA cores and a NVIDIA 4-Plus-1 Quad-Core ARM Cortex-A15 CPU. These embedded chips have been used to power a number of tablets, e.g., Shield Tablet K1.

For the Jetson TK1, we have preloaded all the embedded systems with software supporting the execution of Neurokernel. For information about how to run Neurokernel on the Jetson TK1, see this tutorial. Participants interested in developing fruit fly brain models on Jetson TK1 are encouraged to familiarize themselves with an embedded Linux system through, e.g., elinux.org.