科研 Skill

NeuroMorpho神经元形态检索

neuromorpho-morphology

Finds and compares digital neuron reconstructions on NeuroMorpho.Org by region, cell type or species and reads their morphometry. Use when real neuron morphologies are needed.

发布者 Codar安全审核通过v0.1.0更新于 2026-11-18

SKILL.md

Neuron morphology from NeuroMorpho.Org

NeuroMorpho.Org holds hundreds of thousands of traced neurons, each with metadata and standard morphometry.

Prerequisites

search_datasets and neuron_morphology 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 they are unavailable, use the NeuroMorpho.Org REST API directly, or the "Working with SWC files" section below, which only needs standard Python (or the neurom/navis packages) and works anywhere.

Steps (when brainhound-style search tools are available)

  1. Search with search_datasets, sources: ["neuromorpho"]. Give one term: a brain region (CA1, neocortex), a cell type (pyramidal, Purkinje) or a lab's archive name. Narrow by species (mouse, rat, human). The first line says how many neurons match.
  2. Open a neuron with neuron_morphology, by name (n419) or numeric id. It gives:
    • species, strain, region, cell type, age, sex, stain, protocol and which parts were traced
    • morphometry: total length, surface, volume, stems, bifurcations, branches, size, maximum path distance and branch order, fractal dimension
    • links to the standardized SWC file and a picture
  3. Compare fairly. Check "Reconstructed parts" before comparing lengths: a neuron traced as "Dendrites, Soma, No Axon" can't be compared on axon length. Also check protocol (in vivo vs in vitro), slice thickness and shrinkage correction, which change measured sizes.

Without brainhound: the REST API

import requests

# Search by brain region and species
r = requests.get(
    "https://neuromorpho.org/api/neuron/select",
    params={"q": "brain_region:CA1,species:mouse"},
)
hits = r.json()["_embedded"]["neuronResources"]
print(f"{r.json()['page']['totalElements']} matching neurons")

# Fetch one neuron's metadata by name
meta = requests.get("https://neuromorpho.org/api/neuron/name/n419").json()
print(meta["brain_region"], meta["cell_type"], meta["total_length"])

Working with SWC files

SWC is a plain text format: one point per line with id type x y z radius parent. Type 1 is soma, 2 axon, 3 basal dendrite, 4 apical dendrite. The "standardized" (CNG) version has its soma at the origin and consistent units (µm).

# Total dendritic length from an SWC file, no libraries needed.
import math, urllib.request

url = "https://neuromorpho.org/dableFiles/turner/CNG%20version/n419.CNG.swc"
points = {}
for line in urllib.request.urlopen(url).read().decode().splitlines():
    if line.strip() and not line.startswith("#"):
        i, t, x, y, z, r, parent = line.split()[:7]
        points[int(i)] = (int(t), float(x), float(y), float(z), int(parent))

length = sum(
    math.dist(p[1:4], points[p[4]][1:4])
    for p in points.values()
    if p[4] in points and p[0] in (3, 4)
)
print(f"dendritic length: {length:.0f} µm")

For full analysis, the NeuroM and navis Python packages read SWC files directly.

Limitations

  • Metadata quality (age, protocol, shrinkage correction) varies by contributing lab; cross-dataset morphometry comparisons should check these fields rather than assume consistency.
  • The REST API and SWC downloads require network access to neuromorpho.org.

Licenses and credit

NeuroMorpho.Org content is CC BY 4.0. Its terms ask for all three of these whenever its material is used:

  1. the original paper that describes the reconstructions (the neuron's metadata gives its DOI or PubMed id)
  2. NeuroMorpho.Org (RRID:SCR_002145)
  3. Tecuatl C, Ljungquist B, Ascoli GA (2024) Accelerating the continuous community sharing of digital neuromorphology data. FASEB Bioadv. 6(7):207-221.

Pass all three on unchanged whenever a result is used.

示例提示词

帮我在NeuroMorpho上找这个脑区的锥体神经元重建数据并比较树突形态。

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

NeuroMorpho神经元形态检索能做什么?

Finds and compares digital neuron reconstructions on NeuroMorpho.Org by region, cell type or species and reads their morphometry. Use when real neuron morphologies are needed。

如何安装NeuroMorpho神经元形态检索?

在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「NeuroMorpho神经元形态检索」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。

NeuroMorpho神经元形态检索需要哪些权限?

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。

NeuroMorpho神经元形态检索经过官方审核吗?

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。