Supplementary Code

Plotting scripts and analysis code accompanying the manuscript on Growth-adaptive spring electronics for long-term, same-neuron mapping in the developing rat brain.

---

Overview

We tracked neural populations in neonatal rats (P10-P45) across chronic recording sessions to characterize how individual neurons transition between population coupling states during early postnatal development. Neurons were classified into three developmental trajectory types using Gaussian Mixture Model (GMM) clustering:

• Stable Soloist - consistently low population coupling
• Stable Chorister - consistently high population coupling
• Chorister-to-Soloist - transitions from high to low coupling, primarily during P21-P35

Population coupling is defined as the Pearson correlation between an individual neuron's spike train and the summed population firing rate.

---

Extended Data Figures

Script	Figure	Description
eis.py	ED3	Electrode impedance spectroscopy (Bode plots, incubation, conditions)
individual_neuron_trajectories.py	ED5 a-b	Burst index trajectories for two example units
population_level_trajectories.py	ED5 c-i	Population-level metric trajectories and correlation matrices
cluster_validation_metrics.py	ED6	GMM validation (optimal k, silhouette, AUROC, separation)
per_animal_cluster_normality.py	ED7	Per-animal cluster normality and residual analysis
trajectory_characterisation.py	ED8	Trajectory statistics (age span, timing, variance, effect sizes)
regional_clustering.py	ED9	Regional comparison of V1 and mPFC
metrics_heatmap_lda.py	ED10	14-metric z-scored heatmaps and LDA projections

Main Figure 4 Panels

Script	Panels	Description
gmm_fitting.py	d-e	GMM k=2 clustering with fitted Gaussian components
example_trajectories.py	f-h	Example trajectories for one neuron per trajectory class
populationcoupling_trajectories.py	i-k	Population coupling trajectories split by brain region
population_coupling_lda_visual.py	l	14-metric LDA projection across 8 developmental ages

---

VLM Neuron Tracking Pipeline

The vlm_neuron_tracking/ directory contains a Vision Language Model pipeline for automated neuron matching across chronic recording sessions. The pipeline uses multimodal LLMs (Claude 4, GPT-5.2, and others) to compare waveform morphology, spike location, and firing rate between sessions.

See vlm_neuron_tracking/README.md for full documentation and usage instructions.

---

Requirements

numpy
pandas
matplotlib
scipy

Install with:

pip install numpy pandas matplotlib scipy

The VLM neuron tracking pipeline has additional dependencies listed in vlm_neuron_tracking/requirements.txt.