DL4Proteins Notebook Series provides practical and hands-on resources integrating foundational machine learning concepts with advanced protein engineering methods for predicting and designing protein structures lines.
The series offers accessible learning tools, ranging from neural networks to graph models, that enable researchers, educators, and students to apply deep learning techniques to protein design tasks lines.
The notebooks include introductions to tools like AlphaFold, RFDiffusion, and ProteinMPNN aimed at fostering innovation in synthetic biology and therapeutics lines.
Notebook 1 and Notebook 2 introduce the foundational concepts of neural networks using NumPy and PyTorch, respectively lines.
Notebook 3 explains the foundational concepts of CNNs and demonstrates their application in handling image like data lines.
Notebook 4 explores the use of LMs in understanding sequences such as text and proteins lines.
Notebook 5 delves into the application of language model embeddings in solving real-world problems by repurposing embeddings generated from pre-trained language models lines.
Notebook 6 introduces the use of GNNs in protein research, emphasizing their ability to model the complex relationships between amino acids in protein structures lines.
Notebook 7 explores the application of diffusion models in protein structure prediction and design lines.
Notebook 8 combines advanced tools like RFdiffusion, ProteinMPNN, and AlphaFold to guide users through the complete protein design process lines.