ELLA: Modeling the Subcellular Spatial Variation of Gene Expression within Cells in High-Resolution Spatial Transcriptomics

Usage

Check out the tutorial pages for demos and documentations to get started with ELLA.

Installation

Prerequisites

• https://python-poetry.org/docs/#installation

git clone https://github.com/jadexq/ELLA.git
conda create -n ella python=3.9
conda activate ella
poetry install

Commands

1. Simulate data

   python -m ella.data.simulate_data

   It will generate a simulated_data.json in the current directory
2. Prepare data from pickle

   python -m ella.data.prepare_data -i your_dir/data.pkl -o prepared_data

3. Visualize data

   ella-visualize -d . -c '2102_686' -g 'Alb'

4. Train a model with a recipe, e.g. configs/mini_demo.yaml

   ella-train --config-name mini_demo

   It will save all outputs in ${log.save_dir} (stopping rules is current hard coded!)
5. Open tensorboard for checking convergence

   tensorboard --logdir lightning_logs/debug/gene_0-kernel_-1

6. Estimate

   ella-estimate -d lightning_logs/debug -p "gene_0-kernel_.*" -b 10 -o path/to/out.json

Repo Structure

./ELLA/
├── poetry.lock % poetry dependency lock file
├── pyproject.toml % project metadata and dependencies
├── ella % ELLA source code
│   ├── cli
│   ├── data
│   ├── models
│   ├── utils
│   ├── options.py
│   └── __init__.py
├── docs % source code of the tutorial website
│   └── ...
├── scripts
│   └── demo % code and data for the minimum and complete demos
│       ├── mini_demo
│       └── complete_demo
└── README.md

Processed Data

Complete Demo

• complete_demo_data.pkl

Seq-Scope

• data_health_PC
• data_health_PP
• data_TD_PC
• data_TD_PP

Stereo-seq

• data_Myoblasts.pkl
• data_Cardiomyocytes.pkl

SeqFISH+

• data_Fibroblast.pkl

Merfish mouse brain

• data_EX.pkl
• data_IN.pkl
• data_Astr.pkl
• data_Oligo.pkl