Protein folding with LLM in lighting speed

FoldX focuses on the intersection of biochemistry and artificial intelligence, particularly in the field of protein folding predictions. FoldX utilizes the ESM AI model, which is pre-trained on protein structure data, to predict protein folding structures from primary amino acid sequences. A key component of FoldX's approach is its integration of human feedback. Through a user-friendly interface linked to the PyMOL molecular visualization system, users can interact with the AI-predicted protein structures and offer their feedback. This human feedback is used to refine the predictions of the ESM model and improve its performance over time. The broad range of potential applications includes optimizing engineered bacteria for energy production and accelerating the drug discovery process.

How it works

How to use FoldX with PyMOL

Team

images/team-hanson.png

Hanson

  • A 9-year-old young scientist who began unraveling the complex world of biochemistry before turning 5, showcasing a steadfast commitment to using science as a tool to safeguard our planet, its people, and all creatures that inhabit it.
  • An indefatigable enthusiast possessing a strong passion for translating complex scientific concepts into accessible knowledge for all.

images/team-horace.png

Ace

  • A 7-year-old inventive maker who enjoys crafting intricate creations from everyday items, showcasing an extraordinary blend of patience, persistence, and resourcefulness.
  • An empathetic team-builder with a natural ability to foster connections and uplift those around him.

images/team-wu.png

Wu

  • A patient advisor and fervent advocate fostering engagement in innovative science and engineering projects among the younger generation.
  • An outdoor enthusiast who fervently believes in the value of diverse experiences and rejects the monotony of identical days.

Resource

PyMOL

Presentations