This repository is a curated list of deep learning implementations and resources applied to various biological domains, primarily focusing on genomics. It serves researchers and practitioners by cataloging tools for sequence modeling, protein structure prediction, gene expression analysis, and more, aiming to accelerate the adoption of AI in biological research.

How It Works

The project acts as a comprehensive index, categorizing and linking to numerous open-source projects, research papers, and model repositories. It highlights the application of diverse deep learning architectures, including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Transformers, and Generative Adversarial Networks (GANs), to biological problems. The emphasis is on practical implementations and their underlying methodologies, providing a valuable overview of the state-of-the-art.

Quick Start & Requirements

• Installation and usage vary significantly by the linked project. Many projects are Python-based, often requiring TensorFlow or PyTorch.
• Specific hardware requirements (e.g., GPUs, CUDA) and dependencies (e.g., specific Python versions, large datasets) are project-dependent.
• Links to official documentation, GitHub repositories, and papers are provided for each entry.

Highlighted Details

• Extensive coverage of protein language models (e.g., ESM-1, UniRep, ProGen2) for sequence analysis and structure prediction.
• Detailed sections on genomics, including variant calling (DeepVariant), enhancer prediction, and gene expression analysis.
• Includes cutting-edge applications like AlphaFold for protein structure prediction and transformer models for genomic sequence analysis.

Maintenance & Community

• The list is actively curated, with contributions encouraged.
• Links to related "awesome" lists and specific project communities (e.g., GitHub, Discord) are often provided.

Licensing & Compatibility

• Licenses vary by the individual projects linked. Many are permissive (MIT, Apache), while others may have more restrictive licenses.
• Compatibility for commercial use depends on the specific license of each referenced tool.

Limitations & Caveats

• This is a curated list, not a single software package; users must evaluate and integrate individual tools.
• The rapid pace of deep learning research means some linked resources may become outdated.