neuromotifs 0.1.0a0

Geometry-aware network motif analysis for neocortical microcircuits

neuromotifs

Python tools to load neuronal microcircuit geometry, generate geometry-aware null models, and quantify over/under-expression of 3-node motifs.

Paper: Neuron Morphological Asymmetry Explains Fundamental Network Stereotypy Across Neocortex (Gal et al.)

Install

pip install neuromotifs
# or, for dev
pip install -e .[dev]

Highlights

• Motif counting for directed triplets (#1-#13)
• Geometry-driven random graph generators (1st-5th order) mirroring the paper’s models
• Reproducibility notebooks for Figures 1-4
• Simple CLI: neuromotifs motifs, neuromotifs generate, neuromotifs fit

Quickstart

import pandas as pd
from neuromotifs import MotifCounter, GeometricGenerator

edges = pd.read_csv('data/sample/l5_ttcps_edges.csv')      # u,v directed
pos   = pd.read_csv('data/sample/l5_ttcps_positions.csv')  # id,x,y,z

counter = MotifCounter.from_edges(edges)
counts = counter.count_triplets()
print(counts)

G = GeometricGenerator.from_positions(pos)
G2 = G.generate(order=3, p_mean=0.025)

Data

• data/sample/ contains tiny demonstrators only.
• For full datasets, see data/README.md for scripted download instructions.

Citing

Please cite the paper and this package (see CITATION.cff).