System and method for protein selection

1. A method, comprising:

for each protein mixture in a group of protein mixtures:

for each protein in the protein mixture:

determining an amino acid sequence; and

extracting a feature value vector from the amino acid sequence;

for the protein mixture:

determining a set of amino acid feature values based on the feature value vectors and relative proportions of proteins within the protein mixture;

determining a set of context feature values based on manufacturing specifications for manufacturing a sample using the protein mixture; and

predicting a functional property value for the protein mixture by inputting the set of amino acid feature values and the set of context feature values into a trained prediction model, wherein the trained prediction model is trained using training data comprising, for each training protein mixture in a group of training protein mixtures: a measured functional property value, amino acid sequences, and manufacturing specifications;

selecting a protein mixture from the group based on comparisons between a target functional property value measured for a dairy protein mixture and the predicted functional property value for each protein mixture; and

manufacturing a dairy analog food product, wherein each protein in the selected protein mixture is extracted from a plant source, and wherein the extracted proteins are used as ingredients in the dairy analog food product.

2. The method of claim 1 , further comprising predicting the relative proportions of proteins within the protein mixture based on a source from which the protein mixture is derived.

3. The method of claim 1 , wherein determining the set of amino acid feature values comprises aggregating the feature value vectors based on the relative proportions of the proteins within the protein mixture.

4. The method of claim 3 , wherein aggregating the feature value vectors comprises weighting the feature value vectors based on the relative proportions of the proteins within the protein mixture.

5. The method of claim 3 , wherein aggregating the feature value vectors comprises using a feature aggregation model, wherein the feature aggregation model is a trained machine learning model.

6. The method of claim 5 , wherein the feature aggregation model is trained using multiple instance learning.

7. The method of claim 1 , wherein the manufacturing specifications for the protein mixture comprise at least one of temperature, salt level, pH level, macronutrient ingredients, or microbial ingredients.

8. The method of claim 1 , wherein the set of context feature values is further determined based on modifications to a protein in the protein mixture.

9. The method of claim 8 , wherein the modifications comprise at least one of glycosylation, glycation, phosphorylation, or acylation.

10. The method of claim 1 , wherein extracting the feature value vector comprises determining the feature value vector based on k-mers associated with the amino acid sequence.

11. The method of claim 1 , wherein the feature value vector comprises values for at least one of: pseudo structure status composition (PseSSC), pseudo amino acid composition (PseAAC), or composition, transition, and distribution (CTD).

12. The method of claim 1 , wherein the predicted functional property value comprises a value for at least one of: texture, melt, flavor, chemical properties, denaturation point, particle size, interactions with molecules, or protein aggregation.

13. The method of claim 1 , further comprising selecting the group of protein mixtures based on a set of candidate protein sources, wherein each protein mixture in the group of protein mixtures corresponds to a candidate protein source.

14. The method of claim 1 , wherein the selected protein mixture is further selected based a second predicted functional property value for each protein mixture.

15. The method of claim 1 , wherein manufacturing the dairy analog food product comprises gelling the extracted proteins.

16. The method of claim 1 , wherein the measured function property value comprises at least one of a melt measurement or a texture measurement.

17. The method of claim 1 , wherein the training data comprises, for each training protein mixture in the group of training protein mixtures:

training amino acid feature values determined based on the amino acid sequences for the training protein mixture; and

training context feature values determined based on the manufacturing specifications for the training protein mixture;

wherein the training amino acid feature values and the training context feature values are labeled with the measured functional property value.

18. The method of claim 1 , wherein each protein mixture comprises a subset of proteins in a source, wherein the subset of proteins is selected based on a concentration of each protein in the source.