Computational Protein Prediction for Alternative Food Functions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The widespread adoption of alternative proteins is hindered by high costs and inferior taste compared to conventional meat products, and there is a lack of effective methods for predicting and evaluating proteins with target food functions related to their physicochemical properties.
Innovation Solution
A method involving computer processing of amino acid sequences using a trained model to predict target food functions by identifying intrinsic disorder and determining physicochemical properties, which includes sequence length, molecular weight, and distribution of amino acids, to generate outputs indicative of the protein's functionality, and potentially recombinantly expressing candidate proteins for validation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If alternative proteins are developed to replace animal protein, then environmental sustainability and resource efficiency are improved, but production costs increase and taste quality deteriorates
Solution Approach 1:
The patent applies preliminary action by using trained computer models to predict target food functions and intrinsic disorder in candidate proteins before actual production. This allows for the selection of proteins with desired functional properties (such as meat-like texture and taste) prior to scaling up production, thereby avoiding costly trial-and-error processes and reducing overall production costs while maintaining environmental sustainability benefits
Solution Approach 2:
The patent utilizes parameter changes by analyzing physicochemical properties (such as amino acid composition, molecular weight, and intrinsic disorder parameters) to identify proteins that can replicate the functional characteristics of animal proteins. By optimizing these parameters through computational prediction, the invention enables production of alternative proteins with improved taste and functional properties without increasing production complexity or cost
2Object-affected harmful factors
If alternative proteins are developed to replace animal protein, then environmental sustainability is improved, but taste quality deteriorates
Solution Approach 1:
The patent replaces traditional trial-and-error experimental methods with computational prediction models. The trained computer models analyze amino acid sequences and predict food functions (such as texture, taste, and functional properties) in silico, allowing for the identification of proteins with meat-like sensory characteristics before physical production. This substitution enables precise targeting of taste quality parameters without the resource-intensive experimental process
Solution Approach 2:
The invention focuses on specific physicochemical parameters including intrinsic disorder (IDP content), amino acid composition, and molecular weight that are known to influence protein functionality and sensory properties. By computationally optimizing these parameters to match those of animal proteins, the patent enables development of alternative proteins with improved taste and functional qualities that closely replicate conventional meat products
3Productivity
If computational prediction methods are used to evaluate candidate proteins, then development efficiency is improved, but method complexity increases
Solution Approach 1:
The patent employs a universal computational framework that can evaluate multiple candidate proteins simultaneously using the same trained models. The system analyzes various physicochemical parameters (amino acid composition, molecular weight, intrinsic disorder) through a single integrated prediction platform, enabling high-throughput screening of diverse protein candidates. This multi-functional approach improves development efficiency by consolidating multiple evaluation tasks into one system while managing complexity through standardized algorithms
Data Source
AI summary
A method of predicting at least one target food function of a candidate protein comprises providing an amino acid sequence for the candidate protein; computer processing, by at least one processor executing a trained computer model, the amino acid sequence of the candidate protein to predict a set of candidate amino acid sequences having intrinsic disorder, wherein the intrinsic disorder may comprise a lack of stable secondary structure along at least about 10% of the length of an individual candidate amino acid sequence; and computer processing, by at least one processor executing a trained computer model, the set of candidate amino acid sequences to generate an output that may be indicative of the predicted target food function or target food functions of the candidate protein. Compositions and food products comprising the candidate protein are also disclosed.