# How to Predict Protein Shapes Better Using Lab Tests and Computers > 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. - **Patent:** US 20210174903 - **Original title:** Enhanced protein structure prediction using protein homolog discovery and constrained distograms - **Owner:** Protein Evolution - **Status:** Active - **Times cited:** 21 - **Field:** biotech, software, ai_ml, pharmaceutical, materials ## What it does The patent outlines a method for more accurately predicting the three-dimensional structure of a protein. First, it generates a "multiple sequence alignment" (MSA) of related proteins (claim 1(i)). A computer then uses this MSA to make an initial guess of the protein's shape, creating a "distogram" which is like a map of distances between amino acids (claim 1(ii)). The method then identifies specific amino acids on the protein's surface (claim 1(iii)) and experimentally measures the actual distance between them in a lab, often using a technique called FRET (claim 1(iv), claim 3). These real-world distance measurements are then fed back into the computer model to refine and improve the initial prediction, making it more accurate (claim 1(v)). For example, if the computer initially predicts two surface amino acids are 5 nanometers apart, but the FRET experiment shows they are actually 3 nanometers apart, the computer model is adjusted to reflect the 3 nanometer distance. ## What it does NOT cover - Does not cover protein structure prediction methods that rely solely on computational techniques without any in vitro experimental validation. - Does not cover in vitro distance measurements for amino acids that are not solvent-exposed. - Does not cover methods where the experimental distance measurements are not used to constrain or refine the computational prediction algorithm. - Does not cover protein structure prediction that does not involve generating a multiple sequence alignment (MSA) of homologous sequences. - Does not cover methods that do not produce a distogram as an intermediate step in the structure prediction. ## The clever bit The core innovation is the feedback loop: taking an initial computer prediction, using it to guide specific, targeted lab experiments on solvent-exposed parts of the protein, and then using those precise experimental measurements to correct and improve the computer model. This hybrid approach leverages the strengths of both computational and experimental methods. ## Real-world examples 1. Designing new enzymes for industrial processes. 2. Developing antibodies for therapeutic treatments. 3. Engineering proteins with enhanced stability or function. 4. Predicting the structure of viral proteins for vaccine development. 5. Improving the accuracy of protein structure databases like AlphaFold's predictions. ## Why it matters Understanding a protein's 3D structure is fundamental to biology and drug discovery, as a protein's shape dictates its function. This patent offers a way to make these predictions more reliable by combining the speed of computation with the accuracy of experimental data. More accurate protein structures can accelerate the design of new drugs, enzymes, and materials, impacting fields from medicine to industrial biotechnology. Protein Evolution Inc., the assignee, works in this space, aiming to create new proteins for various applications. ## Frequently asked questions ### What does How to Predict Protein Shapes Better Using Lab Tests and Computers cover? 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. ### Who owns patent US 20210174903? This patent is owned by Protein Evolution. ### When does this patent expire? This patent is expected to expire on December 10, 2040, when the invention enters the public domain. ### What is patent US 20210174903 cited by? This patent has been cited by 21 later patents that build on its ideas. ### What problem does this patent solve? Understanding a protein's 3D structure is fundamental to biology and drug discovery, as a protein's shape dictates its function. This patent offers a way to make these predictions more reliable by combining the speed of computation with the accuracy of experimental data. More accurate protein structures can accelerate the design of new drugs, enzymes, and materials, impacting fields from medicine to industrial biotechnology. Protein Evolution Inc., the assignee, works in this space, aiming to create new proteins for various applications. ### What does this patent NOT cover? Does not cover protein structure prediction methods that rely solely on computational techniques without any in vitro experimental validation. **Full plain-English explainer:** https://patentbrief.org/patent/us/20210174903/enhanced-protein-structure-prediction-using-protein-homolog-discovery-and-constr **Original patent:** https://patents.google.com/patent/US20210174903 --- _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: - [How Computers Predict Protein Shapes Faster Using Smart Templates](https://patentbrief.org/patent/us/20210280268/protein-structure-prediction-system) — This patent describes a computer method to quickly predict the 3D shape of proteins by creating and refining "synthetic templates" from existing protein structures, reducing the heavy computational work usually needed. - [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. - [How Computers Find Hidden Connections Between Different Fields of Knowledge](https://patentbrief.org/patent/us/6523026/google-search-query-processing) — A method for finding related ideas in completely different subjects by using math to map how words appear together, even when the subjects use different vocabulary. - [How Computers Find Similar Text Using Compact Data Structures](https://patentbrief.org/patent/us/10878335/bert-bidirectional-encoder-representations) — This patent describes a method for efficiently identifying similar text records, like documents or product reviews, by using special compact data structures that store text terms probabilistically and then analyzing them with machine learning. - [Using Non-AI Systems to Improve AI Text Generation](https://patentbrief.org/patent/us/12353827/computer-implemented-methods-for-the-automated-analysis-or-use-of-data-including) — This patent describes a method where a traditional, rule-based computer system helps a Large Language Model (LLM) generate more accurate and reliable text by providing it with better context and fact-checking.