Skip to main content
🔍

Installation Guide

Hardware Requirements​

Statescope can operate on a variety of systems depending on the data size:

  • Small-scale data: Suitable for Binder (1 CPU, 2GB RAM on Google Cloud).
  • Large-scale data: Performance improves significantly with higher computational power. Our recommendation for optimal performance is:
    • CPU: 40 threads (Xeon 2.60GHz or equivalent)
    • Memory: 128 GB RAM

Statescope’s Empirical Bayes procedure includes independent optimization steps that can be parallelized to leverage multiple CPUs for faster execution.

Note: For faster performance, the Statescope deconvolution module has GPU support, which can make computations 3-4 times faster.

OS Requirements​

Statescope is primarily tested on Linux-based operating systems. We recommend:

  • Linux distributions: CentOS 7 or Ubuntu 16.04.

Note: While BLADE might work on other OS platforms (macOS, Windows with WSL), these setups may require additional configurations.

Installation Tutorials​

Python Installation​

To manage dependencies, use a Conda enviroment for a consistent environment setup:

  1. Create the Conda environment:

    conda create -n Statescope python=3.10
    conda activate Statescope

  2. Install Statescope

To install Statescope using pip, run the following command:

pip install Statescope

Running a Demo Script​

You can find a basic demo script under the Tutorial. You can open the script using the command below after installing Statescope:

jupyter notebook tutorial/BasicTutorial.ipynb

Additional Information​

Installing Conda​

If you don’t already have Conda installed, you can download and install Miniconda by following these steps:

  1. Go to the Miniconda download page.
  2. Download the appropriate installer for your OS.
  3. Follow the installation instructions for your operating system to set up Conda.

After installation, you can verify Conda by running:

conda --version