# How AI Designs New Protein Sequences for Specific Shapes

> This patent describes a method where a computer-powered neural network fills in missing parts of a protein's amino acid sequence, ensuring the final protein has a desired 3D shape or "fold structure."

- **Patent:** US 12655176
- **Original title:** System and method for generating a protein sequence
- **Owner:** University of Toronto
- **Granted:** 2026
- **Status:** Active
- **Times cited:** 1
- **Field:** biotech, software, ai_ml, pharmaceutical, materials

## What it does

The patent describes a computer-based method for designing protein sequences. It starts by taking an incomplete list of amino acids (called "node elements") for a protein and a map of how these amino acids should physically interact to form a specific 3D shape (an "edge index" defining a "desired fold structure"). A neural network then processes this information. It first figures out improved details about the interactions between amino acids ("enhanced edge attribute values") based on the known amino acids. Then, it uses these improved interaction details to determine better amino acid characteristics ("enhanced amino acid values") for each position. Finally, for any amino acid positions that were initially missing, the network generates the specific amino acid that should go there, completing the protein sequence to match the desired shape. For example, if you wanted a protein with a specific binding pocket, you could provide the known amino acids around that pocket and the desired distances between them, and the network would fill in the rest of the sequence.

## What it does NOT cover

- Does not cover generating protein sequences without defining a desired physical interaction or "fold structure" beforehand.
- Does not cover methods that predict a protein's 3D structure from an existing amino acid sequence.
- Does not cover protein design approaches that do not use a computer-implemented neural network.
- Does not cover generating protein sequences without an iterative refinement process of edge and node attributes.
- Does not cover designing proteins where the "edge attribute value set" does not represent a characteristic of physical interaction, such as a Euclidean distance.

## The clever bit

The novelty lies in the iterative refinement process where the neural network simultaneously enhances both the understanding of physical interactions (edge attributes) and the characteristics of individual amino acids (node attributes) based on a desired 3D structure. This allows it to fill in missing amino acids while ensuring the generated sequence is consistent with the target fold.

## Real-world examples

1. Designing novel enzymes for industrial processes
2. Creating new therapeutic proteins with specific binding sites
3. Developing custom antibodies
4. Engineering proteins for biosensors
5. Designing stable protein scaffolds for drug delivery

## Why it matters

Proteins are the workhorses of biology, performing nearly all cellular functions. Being able to design new protein sequences that reliably fold into specific, desired 3D shapes is crucial for developing new drugs, enzymes, and biomaterials. This technology could accelerate the discovery of novel therapeutics or industrial catalysts by allowing researchers to computationally explore a vast space of potential protein designs.

## Frequently asked questions

### What does How AI Designs New Protein Sequences for Specific Shapes cover?

This patent describes a method where a computer-powered neural network fills in missing parts of a protein's amino acid sequence, ensuring the final protein has a desired 3D shape or "fold structure."

### Who owns patent US 12655176?

University of Toronto owns this patent, granted in 2026.

### When does this patent expire?

This patent is expected to expire on December 4, 2040, when the invention enters the public domain.

### What is patent US 12655176 cited by?

This patent has been cited by 1 later patents that build on its ideas.

### What problem does this patent solve?

Proteins are the workhorses of biology, performing nearly all cellular functions. Being able to design new protein sequences that reliably fold into specific, desired 3D shapes is crucial for developing new drugs, enzymes, and biomaterials. This technology could accelerate the discovery of novel therapeutics or industrial catalysts by allowing researchers to computationally explore a vast space of potential protein designs.

### What does this patent NOT cover?

Does not cover generating protein sequences without defining a desired physical interaction or "fold structure" beforehand.

**Full plain-English explainer:** https://patentbrief.org/patent/us/12655176/system-and-method-for-generating-a-protein-sequence

**Original patent:** https://patents.google.com/patent/US12655176

---

_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 Design and Predict Protein Structures](https://patentbrief.org/patent/us/12603151/methods-of-designing-and-predicting-proteins) — This patent describes a method for predicting or designing proteins by feeding detailed 3D structural information around a specific protein building block into a specially designed deep learning network.
- [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.
- [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.
- [AI Predicts 3D Protein Shapes from Genetic Code](https://patentbrief.org/patent/us/20240412809/protein-structure-prediction-from-amino-acid-sequences-using-self-attention-neur) — 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.
- [Using AI to Design Drugs by Understanding Molecular Shapes](https://patentbrief.org/patent/us/20190304568/system-and-methods-for-machine-learning-for-drug-design-and-discovery) — This patent describes a computer system that uses advanced math to describe molecule shapes and then feeds that information into artificial intelligence to predict how well potential drugs will work or if they'll be toxic.
