Installing LDSC (Python 3.9+ Branch)

This guide covers the installation of the ldsc39 branch from the CBIIT repository.

1. Prerequisites

Ensure you have the following installed on your system:

• git
• Python 3.9 or higher
• conda

2. Clone the Repository

Clone the specific branch (ldsc39) from the CBIIT GitHub repository:

git clone -b ldsc39 https://github.com/CBIIT/ldsc.git
cd ldsc

3. Create a Virtual Environment

It is highly recommended to use a dedicated environment to avoid dependency conflicts.

Using Conda (Recommended):

conda create --name ldsc39 python=3.9
conda activate ldsc39

4. Install Dependencies

Install the required Python packages directly from the requirements.txt file provided in the repo:

pip install -r requirements.txt
pip install numpy
pip install bitarray
pip install pandas
pip install scipy

5. Verify Installation

Run the help command to ensure the script executes correctly:

python ldsc.py -h

6. Download Reference Data

LDSC requires reference LD scores and HapMap3 SNP lists to run. If you are not using the LDscore cloud web tool and are running this locally, you must download these files:

# Example: Download BBJ_HDLC22 LD Scores (approx 5.4MB)
wget https://ldlink.nih.gov/LDlinkRestWeb/copy_and_download/BBJ_HDLC22.txt
# munge sumstats
python munge_sumstats.py --sumstats BBJ_HDLC22.txt --out BBJ_HDLC22
# manually download ref EAS data from 1000 genomes and uncompress, move to folder with ldsc.py script
https://drive.google.com/file/d/1BtpWx02ON33KfjyCFSdmoWYlMZWImh2f/view

7. Basic Usage Example

Once installed, you can run a basic heritability analysis:

python ldsc.py \
    --h2 BBJ_HDLC22.sumstats.gz \
    --ref-ld-chr eas_ldscores/ \
    --w-ld-chr eas_ldscores/ \
    --out your_analysis_results

Technical Note: If you prefer not to manage these local installations and reference files, the LDscore web tool (integrated into LDlink) provides results through a browser without requiring any local setup.

Docker Installation (Alternative)

From the project root directory (where the dockerfile is located):

Clone the Repository

1. Clone the specific branch (ldsc39) from the CBIIT GitHub repository and build the image:

git clone -b ldsc39 https://github.com/CBIIT/ldsc.git
cd ldsc
docker build -t ldsc39 .

2. Run the container:

docker run -d -it --name ldsc39_container -p 5000:5000 ldsc39

3. Access the API:

curl http://localhost:5000/ldscore

Optional - Run a test inside the container:

docker exec -it ldsc39_container bash
# Example: Download BBJ_HDLC22 LD Scores (approx 5.4MB)
wget https://ldlink.nih.gov/LDlinkRestWeb/copy_and_download/BBJ_HDLC22.txt
# munge sumstats
python munge_sumstats.py --sumstats BBJ_HDLC22.txt --out BBJ_HDLC22
# manually download ref EAS data from 1000 genomes and uncompress, move to folder with ldsc.py script
https://drive.google.com/file/d/1BtpWx02ON33KfjyCFSdmoWYlMZWImh2f/view
python ldsc.py --bfile 22 --l2 --ld-wind-cm 1 --out 22