The Fruit Fly Brain Observatory: From Structure to Function

Nikul H. Ukani, Chung-Heng Yeh, Adam Tomkins, Yiyin Zhou, Dorian Florescu, Carlos Luna Ortiz, Yu-Chi Huang, Cheng-Te Wang, Mehmet Kerem Türkcan, Tingkai Liu, Paul Richmond, Chung‐Chuan Lo, Daniel Coca, Ann‐Shyn Chiang, Aurel A. LazarbioRxiv 2019 · DOI

TL;DR

The Fruit Fly Brain Observatory (FFBO) is a new open-source platform designed to unify and democratize access to the vast and scattered data on the fruit fly brain. It provides a comprehensive ecosystem for researchers to store, integrate, visualize, and analyze data, enabling the construction and emulation of computational models to understand brain function.

What Problem Does This Paper Address?

The study of the fruit fly brain, a key model organism for neuroscience, is hampered by the fragmentation of experimental and modeling data across a global community of researchers. Neurobiologists, computational neuroscientists, and computer scientists each work with different datasets and tools, making it difficult to synthesize knowledge and accelerate our understanding of how the brain works. There's a critical need for a centralized, accessible platform that integrates structural data (cell types, connectome, synaptome) with functional data (activity maps) and provides tools for building and testing computational models.

Methods

The FFBO is built around a modular architecture with three main components:

  • NeuroArch: A graph database that stores and integrates diverse biological and computational data, including cell type maps, connectomes, and synaptomes. It uses an object-graph mapping (OGM) for sophisticated querying and allows for the extraction of subcircuits.
  • Neurokernel: A GPU-enabled engine for emulating fruit fly brain circuits. It uses a programming model based on Local Processing Units (LPUs) and supports parallel execution.
  • Frontend Applications (NeuroNLP and NeuroGFX):
    • NeuroNLP: Provides a web-based interface for querying biological data using natural language. It allows users to explore cell type maps, connectomes, and synaptomes, and access neurophysiology data for activity maps.
  • NeuroGFX: Offers a graphical user interface for building, manipulating, and visualizing executable neural circuit models. It allows users to explore circuit function at different levels of abstraction, from individual neurons to whole brain models.

Datasets

The FFBO integrates existing datasets from various sources, including:

  • FlyCircuit: A large dataset of fruit fly brain connectomics.
  • Janelia Research Campus: Provides EM-based reconstructions of the medulla and other brain regions.
  • Drosophila RNAi Screening Center (DRSC) and Transgenic RNAi Project (TRiP): Sources for functional genomics data.
  • Vienna Drosophila Research Center: Offers RNAi libraries.
  • FlyLight GAL4 and Split-GAL4 lines: Resources for genetic manipulation.
  • DoOR database: For mapping Drosophila odorant responses.
  • Neurodata Without Borders (NWB): A standard format for neurophysiology data, which FFBO supports for data upload and download.

The paper does not introduce new benchmark datasets but rather focuses on integrating and making accessible existing ones.

Key Insights and Ideas

  • Unifying fragmented data: The FFBO's core contribution is creating a single, accessible ecosystem for the diverse data types related to the fruit fly brain, bridging the gap between structural and functional neuroscience.
  • Natural Language Querying: NeuroNLP's natural language interface significantly lowers the barrier to entry for accessing complex neurobiological data, allowing researchers to query information without deep knowledge of query languages.
  • Bridging Structure and Function: The platform directly links structural data (connectome, synaptome) to functional data (activity maps) and computational models, enabling researchers to test hypotheses about how brain structure gives rise to function.
  • Collaborative Exploration: Features like "tags" for sharing session states and an open-source architecture foster collaboration and dissemination of findings within the research community.
  • Scalability and Extensibility: The underlying design is intended to be scalable and extensible, allowing for the integration of new data types and tools as the field advances.

Key Results

The paper demonstrates the FFBO's capabilities through examples:

  • Cell Type Mapping: NeuroNLP can be used to query and visualize specific neurons and their morphologies, aiding in the expansion of existing cell type maps.
  • Connectome Exploration: Users can build neural pathways by querying connectivity data, as shown in the construction of a visual motion detection pathway.
  • Synaptome Visualization: The platform allows for the visualization of synapse distributions, providing insights into the connectivity patterns of specific neurons.
  • Activity Map Access: FFBO provides an interface to search and visualize neurophysiology data, contributing to the construction of an activity map.
  • Circuit Emulation: NeuroGFX enables the construction and execution of computational models of brain circuits, allowing for the exploration of circuit function.

Limitations and Future Work

  • Data Completeness: While the FFBO integrates a vast amount of data, the paper acknowledges that the connectome and synaptome are still incomplete, especially for certain brain regions.
  • Model Format Standardization: The authors note that a standardized model format for neural circuits has not yet been widely adopted in the literature, which could hinder seamless integration of models from different sources.
  • Future Work: The authors envision expanding the FFBO with additional tools, such as an in-silico experimental workbench and support for in-silico optogenetic experiments. They also highlight the ongoing development of FlyBrainLab, an interactive open computing platform that builds upon FFBO's capabilities. The platform is designed to be easily extended with new data types and components.

Summary updated

Abstract

Abstract The fruit fly is a key model organism for studying the activity of interconnected brain circuits. A large scattered global research community of neurobiologists and neurogeneticists, computational and theoretical neuroscientists, and computer scientists and engineers has been developing a vast trove of experimental and modeling data that has yet to be distilled into new knowledge and understanding of the functional logic of the brain. Developing open shared models, modelling tools and data repositories that can be accessed from anywhere in the world is the necessary engine for accelerating our understanding of how the brain works. To that end we developed the Fruit Fly Brain Observatory (FFBO), the next generation open-source platform to support open, collaborative Drosophila neuroscience research. FFBO provides a (i) hub for storing and integrating fruit fly brain research data from multiple data sources worldwide, (ii) unified repository of tools and methods to build, emulate and compare fruit fly brain models in health and disease, and (iii) an open framework for fruit fly brain data processing and model execution. FFBO provides access to application tools for visualizing, configuring, simulating and analyzing computational models of brain circuits of the (i) cell type map, (ii) connectome, (iii) synaptome, and (iv) activity map using intuitive queries in plain English. Tools are provided to extract the function inherent in these structural maps. All applications can be accessed with any modern browser.
BibTeX
@misc{ukani2019fruit,
  author = {Nikul H. Ukani and Chung-Heng Yeh and Adam Tomkins and Yiyin Zhou and Dorian Florescu and Carlos Luna Ortiz and Yu-Chi Huang and Cheng-Te Wang and Mehmet Kerem Türkcan and Tingkai Liu and Paul Richmond and Chung‐Chuan Lo and Daniel Coca and Ann‐Shyn Chiang and Aurel A. Lazar},
  title = {{The Fruit Fly Brain Observatory: From Structure to Function}},
  year = {2019},
  doi = {10.1101/580290},
  howpublished = {bioRxiv},
}
Figure