科研 Skill

Allen脑区本体查询

allen-brain-regions

Looks up Allen Institute brain region names, acronyms and atlas hierarchy, and picks a BrainGlobe reference atlas. Use when matching region labels across datasets or papers.

发布者 Codar安全审核通过v0.1.0更新于 2026-11-18

SKILL.md

Brain regions and atlases

Labs write the same region many ways: "CA1", "field CA1", "hippocampal CA1", "dorsal CA1". The Allen ontology gives each region one acronym, one numeric id and a place in a hierarchy, so data from different sources can be lined up.

Prerequisites

brain_region and brain_atlases below are tools from the upstream "brainhound" agent harness this skill was written for. They are not bundled with this skill and will not exist as callable tools in a generic Codex/Claude Code environment. If those tools are not available, use the "Using an atlas in Python" section instead, which only needs pip install brainglobe-atlasapi and works anywhere.

Steps (when brainhound-style lookup tools are available)

  1. Look up the region with brain_region. Give an acronym (CA1, VISp, MOs) or part of a name (primary visual). It tries an exact acronym first, then names that contain the text. It returns:
    • the acronym, full name and numeric id
    • the color used in Allen's atlas images
    • the chain of parents, such as root > grey > CH > CTX > CTXpl > HPF > HIP > CA > CA1
  2. Pick the species with species: mouse (default, the CCFv3 regions), human, or developing mouse. Human atlas names differ from mouse ones; try adjectives like hippocampal if a noun finds nothing.
  3. Match broadly when needed. A dataset labeled HPF (hippocampal formation) contains CA1, CA3, DG and more. Use the parent chain to decide whether two labels overlap, and say so rather than treating them as the same region.
  4. Choose an atlas with brain_atlases. It lists BrainGlobe's atlases (mouse, rat, human, zebrafish and others) with species, voxel size and citation. Filter with words, like allen mouse or zfish.

Using an atlas in Python (works without any special tools)

# pip install brainglobe-atlasapi
from brainglobe_atlasapi import BrainGlobeAtlas

atlas = BrainGlobeAtlas("allen_mouse_25um")      # downloads the atlas once
ca1 = atlas.structures["CA1"]                      # same acronym and id as brain_region
mask = atlas.get_structure_mask("CA1")             # a 3D boolean volume
print(atlas.structure_from_coords((264, 160, 228), as_acronym=True))  # voxel coordinates

atlas.structures is a dict keyed by acronym with id, name and structure_id_path (the parent chain), so the same region-matching logic from Steps 1-3 above can be done directly against it without any external tool.

Limitations

  • Region matching is name/acronym based; it does not register images or resolve coordinates between atlas spaces. Use a registration tool (e.g. BrainGlobe's brainreg) for that.
  • CCFv3 and other BrainGlobe atlases are mouse/rat/human reference templates; an individual subject's anatomy can deviate from them.

Licenses and credit

  • Allen Institute content may be used for research and other noncommercial purposes only, derivative works included. Flag this if the user's use could be commercial. Cite the CCFv3 paper (Wang et al., 2020, Cell) for mouse regions.
  • Each BrainGlobe atlas has its own citation (atlas.metadata or, with brainhound, brain_atlases). Cite BrainGlobe too (Claudi et al., 2020, JOSS).

示例提示词

帮我用Allen脑图谱查这个脑区的标准名称和缩写对应关系。

所属插件

神经科学面向神经科学研究:EEG/MEG、fMRI、BIDS、spike sorting、连接组、脑机接口解码评估、神经刺激安全与公开神经数据检索。

安装该插件会一并获得同领域的 45 个 Skill / MCP。

同插件的其他科研 Skill

常见问题

Allen脑区本体查询能做什么?

Looks up Allen Institute brain region names, acronyms and atlas hierarchy, and picks a BrainGlobe reference atlas. Use when matching region labels across datasets or papers。

如何安装Allen脑区本体查询?

在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「Allen脑区本体查询」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。

Allen脑区本体查询需要哪些权限?

该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。

Allen脑区本体查询经过官方审核吗?

是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。