Set up complete project structure for single-cell and spatial transcriptomics analysis

[Copilot]

This PR implements the complete project structure for organizing single-cell RNA-seq and spatial transcriptomics analysis projects focused on thoracic oncology.

Changes Overview

Directory Structure

Created three main directories with comprehensive README documentation:

• raw_data/ - For storing original datasets from 10x Genomics, GEO/GSE, and Visium platforms. The README provides guidance on organizing data by source type.

• notebooks/ - For Jupyter notebooks containing analysis workflows. The README documents the types of analyses supported (preprocessing, clustering, trajectory analysis, spatial deconvolution) and the main tools used.

• figures/ - For storing generated plots and visualizations (UMAP projections, dot plots, heatmaps, spatial plots, etc.). The README recommends organization strategies and output formats.

Environment Configuration (environment.yml)

Created a comprehensive conda environment specification with all dependencies required for single-cell and spatial transcriptomics analysis:

Core Analysis Libraries:

• Scanpy (>=1.9.3) - Primary single-cell analysis framework
• AnnData (>=0.9.0) - Annotated data structures
• CellRank (>=2.0.0) - Trajectory and cell fate analysis
• scVelo - RNA velocity analysis
• Squidpy (>=1.2.0) - Spatial transcriptomics tools

Dimensionality Reduction & Clustering:

• umap-learn (>=0.5.3)
• leidenalg (>=0.9.1)
• louvain (>=0.8.0)

Supporting Libraries:

• Data manipulation: pandas, numpy, scipy, scikit-learn
• Visualization: matplotlib, seaborn, plotly
• Notebook environment: jupyter, jupyterlab, ipykernel, ipywidgets
• File I/O: h5py, tables, xlrd
• Additional tools: gseapy, session-info

Git Ignore Configuration (.gitignore)

Implemented comprehensive ignore patterns to exclude:

• Large data files in raw_data/ (h5, h5ad, loom, mtx, compressed archives)
• Python cache files and bytecode (__pycache__, *.pyc)
• Jupyter notebook checkpoints (.ipynb_checkpoints)
• Virtual environments and IDE configuration files
• OS-specific files (.DS_Store, Thumbs.db)
• Scanpy cache directories

The .gitignore preserves README files in data directories while excluding large binary data files.

Usage

To set up the environment:

conda env create -f environment.yml
conda activate singlecell-projects

This structure provides a clean foundation for reproducible single-cell and spatial transcriptomics analysis workflows, with clear separation of data, code, and outputs.

Original prompt

singlecell-projects/
├── raw_data/ # Dados originais (10x, GSE, Visium)
├── notebooks/ # Notebooks de análise e visualização
├── figures/ # Plots e imagens de UMAP, clusters etc.
├── environment.yml # Dependências de Scanpy, AnnData, UMAP, CellRank, etc.
├── .gitignore
└── README.md

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.