Deep Learning for Life Sciences

Example code from the book "Deep Learning for the Life Sciences". This repository provides hands-on examples applying deep learning to problems such as drug discovery, genomics, and protein structure prediction.

Features

• Example code and tutorials from the book
• Ready-to-use Docker environment with DeepChem 2.4.0
• GPU acceleration support (NVIDIA)

Installation

1. Docker Image (Recommended)

You can use the pre-built Docker image from DockerHub:

• Image: deepchemio/deepchem:2.4.0
  • Built using conda (DeepChem version 2.4.0)
  • Dockerfile is located in the docker/tag directory

Install Docker

Follow the official Docker installation guide.

2. Clone This Repository

Clone the repository to your local machine:

git clone https://github.com/partrita/DeepLearningLifeSciences.git
cd DeepLearningLifeSciences

Usage

1. Build the Docker Image Locally

If you want to build your own image (for example, after modifying code):

docker build -t deeplife .

2. Run the Docker Container

Mount your current directory into the container for easy file access:

docker run --rm -it -v $PWD:/root/mydir/ deeplife

GPU Support

• If nvidia-docker is installed:

  nvidia-docker run --rm -it -v $PWD:/root/mydir/ deeplife

  or

  docker run --runtime nvidia --rm -it -v $PWD:/root/mydir/ deeplife

• If nvidia-container-toolkit is installed:

  docker run --gpus all --rm -it -v $PWD:/root/mydir/ deeplife

3. Using the Official DeepChem Image

You can also use the official DeepChem image directly:

docker run --rm -it -v $PWD:/root/mydir/ deepchemio/deepchem:2.4.0

References

• Deep Learning for the Life Sciences (O'Reilly)
• DeepChem Documentation
• Docker Documentation

License

MIT License. See LICENSE for details.