morph-spines 1.1.0

Package to load and write morphologies with spines

morph-spines

A Python library for loading, writing, and accessing neuron morphologies with dendritic spine data from HDF5 files. It provides structured access to spine skeletons, meshes, and spatial transformations.

Quick example

Loading

from morph_spines import load_morphology_with_spines

m = load_morphology_with_spines("neuron.h5", spines_are_centered=True, load_meshes=True)

# Access spine meshes
mesh = m.spines.spine_mesh(0)
print(mesh.vertices.shape, mesh.faces.shape)

# Get only the head region of a spine
head_mesh = m.spines.spine_mesh(0, include_neck=False)

# Spine type classification
spine_type = m.spines.spine_type(0)

Writing

from morph_spines import write_spine_table, write_morphology, write_soma_mesh

# Write a spine table (pandas DataFrame with mandatory columns)
write_spine_table("output.h5", "neuron_01", spine_table_df)

# Write neuron morphology skeleton
write_morphology("output.h5", "neuron_01", points, structure)

# Write soma mesh
write_soma_mesh("output.h5", "neuron_01", vertices, triangles)

Validation

from morph_spines import validate_morph_with_spines_file

# Check file structure only (groups, datasets, metadata)
result = validate_morph_with_spines_file("neuron.h5")

# Also check data integrity (shapes, dtypes, value ranges, cross-references)
result = validate_morph_with_spines_file("neuron.h5", check_data_integrity=True)

print(result)          # Human-readable summary
assert result.is_valid # Use programmatically

### Merging

```python
from pathlib import Path
from morph_spines import merge_morphologies_with_spines

# Merge multiple files without renaming
merge_morphologies_with_spines(
    source_files=[Path("neuron_A.h5"), Path("neuron_B.h5")],
    output_path=Path("merged.h5"),
)

# Merge with renaming (neuron keys and/or spines library names)
src1 = Path("neuron_A.h5")
src2 = Path("neuron_B.h5")
merge_morphologies_with_spines(
    source_files=[src1, src2],
    output_path=Path("merged.h5"),
    rename_map={
        (src1, "morph_001"): "circuit_neuron_42",
        (src2, "morph_001"): "circuit_neuron_43",
    },
)

# Merge without meshes (smaller output)
merge_morphologies_with_spines(
    source_files=[Path("a.h5"), Path("b.h5")],
    output_path=Path("merged_no_meshes.h5"),
    include_meshes=False,
)

Installation

pip install morph-spines

For development:

git clone https://github.com/openbraininstitute/morph-spines.git
cd morph-spines
pip install -e ".[test]"

Features

• Load and write neuron morphologies with spine data from/to HDF5 files
• Merge multiple morph-with-spines files into one, with optional renaming of neuron keys and spines library names
• Access the spine table with per-spine properties (position, orientation, section placement)
• Access spine skeletons (via NeuroM/MorphIO) and meshes (via trimesh)
• Write spine tables, morphologies, soma meshes, spine meshes, and spine skeletons
• Validate spine tables against the format specification before writing
• Validate entire morph-with-spines files (structure and optionally data integrity)
• Head/neck triangle classification with filtering (include_head, include_neck)
• Support for branched spines with multiple heads
• Spine type classification (thin, mushroom, stubby, filopodium, branched, etc.)
• Lazy or eager mesh loading
• Coordinate transformations between local spine and global neuron frames

Upgrading from v0.x

Version 1.0 drops support for reading spine tables stored as pandas DataFrames (v0.1 format) inside HDF5 files. If you have files in the old format, convert them before loading:

python scripts/h5_dataframe_to_h5_datasets_group.py old_file.h5 new_file.h5

The conversion script requires the tables package:

pip install morph-spines[scripts]

File format

The morphology-with-spines format is documented in examples/data/README.md.

Development

Run tests:

pytest

Lint:

ruff check src/ tests/

Type check:

mypy src/

Examples

See the examples/ folder for Jupyter notebooks demonstrating visualization and usage.

License

Copyright (c) 2025-2026 Open Brain Institute.

Licensed under Apache-2.0.