HumanMorphoMap

A computational framework for constructing statistical morphology maps of human cortical cells, integrating multi-modal morphological and transcriptomic data.

Overview

This project aims to:

• Create probabilistic morphology atlases of human cortical neurons
• Correlate cellular morphology with spatial transcriptomics
• Enable post-tracing guided morphology refinement based on H01 priors using GMM model

Key Features

✔️ H01-guided neuronal morphology refinement pipeline
✔️ Spatial transcriptomics integration
✔️ Morphological analyses across various conditions

Repository Structure

├── h01-guided-reconstruction/ # H01 dataset-guided morphology refinement
├── meta/ # Project metadata and schemas
├── resources/public_data/ # publicly downloaded morphological datasets
├── spatial_transcript_seu/ # Spatial transcriptomics analysis
├── src/ # Core computational pipelines

Installation and dependency

• First, download the project

git clone https://github.com/SEU-ALLEN-codebase/HumanMorphoMap.git

• install dependencies Major dependencies including: Vaa3D-x (version 1.1.4), scanpy (version 1.11.2), and cell2location (version 0.1.4). PyTorch version 2.71.+cu12.6 was utilized for cell2location.

Usage

See subfolder READMEs for module-specific instructions.

License

Apache-2.0 license