Using NLP to Detect Knots in Protein Structures

Explore the fascinating application of Natural Language Processing (NLP) in detecting knots in protein structures during this 25-minute conference talk from EuroPython 2023. Delve into the world of protein 3D structures and their importance in biological functions. Learn how recent advancements in Machine Learning and NLP have revolutionized protein structure prediction, enabling models like AlphaFold to achieve lab-quality results. Discover the specific use of language models in identifying protein knots solely from amino acid sequences, and understand how this phenomenon affects protein function and stability. Gain insights into the Python-based tools utilized in this research, including Biopython, Pymol, and the Hugging Face transformer library. Uncover the potential of NLP and interpretation techniques in deciphering the mechanisms behind protein knotting, opening new avenues for understanding these essential biological components.