Brain Reference Architecture Editorial System (BRAES)

BRAES is an online platform for the creation, submission, review, and publication of Brain Reference Architecture (BRA) data. Developed by the Whole Brain Architecture Initiative (WBAI), it aims to facilitate the development of brain-inspired software based on brain reverse engineering.

Brain Information Framework (BIF) data: Anatomical data of the target brain region Hypothetical Component Diagram (HCD): Functionality data assigned to each brain region

1. a) Extraction of anatomical knowledge for the Region of Interest (ROI)
2. b) Determination of ROI and Top Level Function (TLF)
3. c) Decomposition of TLF based on ROI's internal structure and determination of computational functions for each uniform circuit

• Provision of manuals and tutorials for BRA data design
• Acceptance of BRA data submissions
• Review process by neuroscience and engineering experts
• Publication of approved BRA data under Creative Commons license

• Authenticity of BIF
• Consistency of HCD with BIF
• Functionality of HCD

BRAES aims to accelerate the elucidation of computational functions of the entire brain by widely sharing BRA data. As data accumulates, it becomes possible to estimate computational functions of unknown uniform circuits, with the expectation that assignment of computational functions will progress rapidly once a critical point is reached. This system serves as a crucial resource for researchers and developers working in the field of brain-inspired computing, supporting WBAI's efforts to realize a whole brain architecture by 2026.

Poster

1. Yoshimasa Tawatsuji, Yuta Ashihara, Naoya Arakawa, Hiroshi Yamakawa, BRAES: Developing Brain Reference Architecture Editorial System for Accumulating Hypotheses on Neural Functionality, INCF Neuroinformatics Assembly 2023. https://training.incf.org/lesson/braes-developing-brain-reference-architecture-editorial-system-accumulating-hypotheses.
2. BRA-driven development

Hello everyone, I'm Yoshimasa Tawatsuji from the University of Tokyo. Today, my talk title is “BRAES: Developing Brain Reference Architecture Editorial System for Accumulating Hypotheses on Neural Functionality.” First of all, we at WBAI (Whole Brain Architecture Initiative) aim to realize a whole brain architecture by 2026. In brain architecture, functions are assigned to major brain organs such as neocortex, nuclei in basal ganglia, and the hippocampus, based on the brain reference architecture approach. The results of such work are summarized as BRA data, which consists of anatomical data on the target brain region and data on the functionality assigned to each brain region. In this presentation, we will introduce BRAES (BRA Editorial System), a system for submitting and reviewing designed BRA data. The design of BRA data consists of three major steps:

1. A contributor extracts knowledge about the anatomy of the region of interest from neuroscience papers and data. This is collected information on neuron groups called uniform circuits, and we call it BIF (Brain Information Framework) data.
2. The contributor determines the ROI (Region of Interest) and the TLF (Top Level Function).
3. Based on the internal structure of the ROI, the contributor decomposes the TLF and determines the computational function of each uniform circuit. We call this Hypothetical Component Diagram (HCD).

BIF data and HCD data are collectively called BRA data. BRA data is designed in Google Spreadsheet format. BRAES is a system for submitting, reviewing, and publishing data with the aim of widely sharing BRA data. There are four main functions:

1. Providing manuals and tutorials for BRA data design
2. Receiving submissions of BRA data
3. Reviewing BRA data
4. Publishing accepted BRA data under a Creative Commons license

The data will be reviewed by neuroscience and engineering experts from three perspectives:

• Authenticity of the BIF
• Consistency of HCD with the BIF
• Functionality of the HCD

We have also designed a manual for the reviewers. We believe that broadly sharing BRA data will contribute to the rapid elucidation of the computational functions of the entire brain. By accumulating BRA data, we can estimate the computational functions of unknown uniform circuits. This estimation becomes easier as more data is accumulated. In other words, the candidates for computational functions to be assigned become more limited, and at some critical point, the assignment of computational functions will proceed rapidly. Please submit BRA data. More details will be discussed in the poster. We look forward to hearing your opinions. Thank you very much.