# AI Predicts 3D Protein Shapes from Genetic Code > This patent describes how a computer uses self-attention neural networks to predict the complex three-dimensional shape of a protein based on its simple amino acid sequence, a process crucial for understanding biology and developing new medicines. - **Patent:** US 20240412809 - **Original title:** Protein Structure Prediction from Amino Acid Sequences Using Self-Attention Neural Networks - **Owner:** DeepMind Technologies - **Status:** Active - **Times cited:** 0 - **Field:** biotech, ai_ml, software, pharmaceutical, telecommunications ## What it does This method, performed by computers, predicts a protein's 3D structure from its amino acid sequence. First, it takes the amino acid sequence and creates an initial 'embedding' for every possible pair of amino acids in that sequence (Claim 1). Optionally, this initial embedding can be improved by comparing the protein's sequence to similar proteins from other species, creating a 'multiple sequence alignment' (Claim 2). Next, these initial pair embeddings are fed into a 'pair embedding neural network' that uses multiple 'self-attention neural network layers' to refine them into 'final embeddings' (Claim 1). Each self-attention layer updates an embedding by looking at other related embeddings, sometimes focusing only on pairs in the same 'row' or 'column' of a 2D arrangement (Claims 3-6). Finally, the computer determines the protein's predicted 3D structure using these refined final embeddings (Claim 1). For example, if you input the sequence for insulin, the system would output its predicted folded shape. ## What it does NOT cover - Does not cover methods that determine protein structure using physical experiments like X-ray crystallography or cryo-electron microscopy. - Does not cover protein structure prediction systems that do not use self-attention neural network layers for processing pair embeddings. - Does not cover methods that predict protein structure without first generating embeddings for pairs of amino acids. - Does not cover systems that predict protein function or interactions without determining the 3D structure. - Does not cover predictions based on inputs other than an amino acid sequence, such as a protein's physical properties or environment. ## The clever bit The innovation lies in using 'self-attention neural networks' specifically on 'pair embeddings' of amino acids. This allows the system to efficiently learn complex relationships between distant parts of a protein, which is crucial for accurately predicting how the entire chain folds into a 3D shape, especially with the alternating row-wise and column-wise attention layers. ## Real-world examples 1. DeepMind's AlphaFold 2. AlphaFold 2 3. Protein structure prediction services 4. Computational drug discovery platforms ## Why it matters Understanding a protein's 3D shape is fundamental to knowing how it works. This technology speeds up scientific discovery in fields like drug design, where knowing a protein's structure helps scientists create molecules that can interact with it. It also aids in understanding diseases caused by misfolded proteins and designing new enzymes for industrial applications. ## Frequently asked questions ### What does AI Predicts 3D Protein Shapes from Genetic Code cover? This patent describes how a computer uses self-attention neural networks to predict the complex three-dimensional shape of a protein based on its simple amino acid sequence, a process crucial for understanding biology and developing new medicines. ### Who owns patent US 20240412809? This patent is owned by DeepMind Technologies. ### When does this patent expire? This patent is expected to expire on August 23, 2044, when the invention enters the public domain. ### What problem does this patent solve? Understanding a protein's 3D shape is fundamental to knowing how it works. This technology speeds up scientific discovery in fields like drug design, where knowing a protein's structure helps scientists create molecules that can interact with it. It also aids in understanding diseases caused by misfolded proteins and designing new enzymes for industrial applications. ### What does this patent NOT cover? Does not cover methods that determine protein structure using physical experiments like X-ray crystallography or cryo-electron microscopy. **Full plain-English explainer:** https://patentbrief.org/patent/us/20240412809/protein-structure-prediction-from-amino-acid-sequences-using-self-attention-neur **Original patent:** https://patents.google.com/patent/US20240412809 --- _Source: PatentBrief — https://patentbrief.org. Patent facts are from public records; the plain-English explanation is PatentBrief's._ ## Related patents Semantically similar inventions in the PatentBrief corpus: - [Using AI to Predict Protein Shapes for Drug Discovery](https://patentbrief.org/patent/us/12437843/predicting-protein-structures-using-geometry-neural-networks-that-estimate-simil) — This patent describes an AI-driven method for iteratively predicting protein 3D structures using a geometry neural network, then using those predictions to find and synthesize new drug molecules. - [Computer Model Designs Proteins by Matching Structure to Sequence](https://patentbrief.org/patent/us/12412637/embedding-based-generative-model-for-protein-design) — IBM's 2025 patent describes a computer system that designs new protein sequences by first understanding a target 3D protein shape and then generating a sequence that would fold into that shape. - [Predicting How Molecules Interact with Proteins Using Two AI Streams](https://patentbrief.org/patent/us/11176462/system-and-method-for-prediction-of-protein-ligand-interactions-and-their-bioact) — This patent describes a computer system that uses two separate artificial intelligence models, one for molecules and one for proteins, to predict how they will interact and what biological effects they might have. - [How to Predict Protein Shapes Better Using Lab Tests and Computers](https://patentbrief.org/patent/us/20210174903/enhanced-protein-structure-prediction-using-protein-homolog-discovery-and-constr) — This patent describes a method to improve predicting a protein's 3D shape by combining computer simulations with actual distance measurements from specific parts of the protein in a lab. - [AI Model Predicts Protein Shapes from Infrared Light Data](https://patentbrief.org/patent/us/10962473/protein-secondary-structure-prediction) — This patent describes a computer method that uses an artificial intelligence model to predict the detailed 3D shapes of proteins within a food ingredient by analyzing how the ingredient absorbs infrared light.