Synapse Subtype Analyzer
A classification engine for identifying synaptic subtypes from multi-protein colocalization data. Processes detected synapse markers (PSD95, SAP102, GluN1, etc.) and classifies each synapse into molecularly defined subtypes based on marker presence and intensity patterns.
Solo Developer
2 weeks
Reader-first case study
This case study focuses on problem framing, implementation choices, technical constraints, and outcome.
Solo Developer
2 weeks
Research project
2024
Where parts of the system are internal or institutional, this case study focuses on engineering scope, workflow design, and technical decisions rather than trying to simulate missing public artifacts.
Project Overview
The Synapse Subtype Analyzer classifies individual synapses into molecularly defined subtypes based on which synaptic proteins are detected at each location. By analysing colocalization patterns across multiple protein markers, it categorises every detected synapse in a brain section — enabling researchers to map the molecular diversity of synapses across brain regions.
Project Details
Technologies Used
- Python 3
- NumPy (numerical processing)
- pandas (data management)
- Classification logic (rule-based and threshold-based)
Key Features
- Multi-Marker Classification: Analyses combinations of synaptic protein markers (PSD95, SAP102, GluN1, GluA1, GluA2, etc.) to determine subtype identity
- Colocalization Logic: Determines which proteins are co-present at each synapse location based on spatial proximity and intensity thresholds
- Subtype Taxonomy: Classifies synapses into defined molecular subtypes (e.g., Type 1: PSD95+/SAP102-, Type 2: PSD95+/SAP102+, etc.)
- Region-Level Statistics: Aggregates subtype distributions per brain region for comparative analysis
- Batch Processing: Processes entire experimental datasets across multiple brain sections and regions
- Pipeline Integration: Feeds directly into downstream heatmap generation and statistical analysis tools
Technical Highlights
- Combinatorial Classification: Handles the combinatorial explosion of possible marker combinations (2^n subtypes for n markers) with efficient lookup-based classification
- Threshold Calibration: Configurable intensity thresholds per marker to account for different antibody efficiencies and imaging conditions
- Output Compatibility: Results formatted for direct input into the brain heatmap generator and statistical analysis pipeline
Impact
Core component of the synaptome mapping pipeline. Subtype distributions generated by this tool have been used in published research to characterise the molecular architecture of synapses across different brain regions, ages, and experimental conditions.