Back to results
US 2026/0260707 A1Published application
Artificial intelligence-mediated methods and systems for genome editing
AssigneePioneer Hi-Bred International, Inc.
InventorELI RODGERS-MELNICK
Life of this patent
Estimating the term…
Abstract
Described herein are artificial intelligence-mediated methods and systems for genome editing in a plant.
Claims55
An artificial intelligence model-mediated method for editing a plant genome, the method comprising:
- (a) providing an artificial intelligence (AI) model with a first dataset, the first dataset comprising a reference nucleotide sequence of at least one plant regulatory element;
- (b) providing the AI with a second dataset, the second dataset comprising a plurality of variant nucleotide sequences of the reference nucleotide sequence;
- (c) predicting one or more expression profiles for each variant nucleotide sequence of the plurality of variant nucleotide sequences relative to expression of the reference nucleotide sequence;
- (d) calculating a first fitness score for each variant nucleotide sequence, wherein the first fitness score reflects the degree to which a predicted expression profile for a variant nucleotide sequence meets a target expression profile, and wherein the first fitness score incorporates one or more constraints that alter the suitability of the variant nucleotide sequence based on a genome editing system;
- (e) selecting at least one final variant nucleotide sequence from the plurality of variant nucleotide sequences;
- and
- (f) editing the plant genome.
4 more elements
An artificial intelligence method for predicting expression modifications due to genetic variants, the method comprising:
- (a) providing an artificial intelligence (AI) model with a first dataset, the first dataset comprising a reference nucleotide sequence of at least one plant regulatory element;
- (b) providing the AI model with a second dataset, the second dataset comprising a plurality of variant nucleotide sequences of the reference nucleotide sequence;
- (c) predicting one or more expression profiles of each variant nucleotide sequence of the plurality of variant nucleotide sequences relative to expression of the reference nucleotide sequence;
- and
- (d) calculating a first fitness score for each variant nucleotide sequence, wherein the first fitness score reflects the degree to which a predicted expression profile for a variant nucleotide sequence meets a target expression profile, and wherein the first fitness score incorporates one or more constraints that alter the suitability of the variant nucleotide sequence.
2 more elements
An artificial intelligence model-mediated method for breeding genetically modified plants, the method comprising:
- calculating a fitness score for each of a plurality of variant nucleotide sequences of a plant regulatory element, wherein calculating the fitness score comprises providing an artificial intelligence (AI) model with a first dataset comprising a reference nucleotide sequence of the plant regulatory element and a second dataset comprising the plurality of variant nucleotide sequences of the plant regulatory element and predicting one or more expression profiles of each of the plurality of variant nucleotide sequences relative to expression of the reference nucleotide sequence;
- selecting a variant nucleotide sequence from the plurality of variant nucleotide sequences based on the fitness score;
- providing a plant cell with a genome editing system that edits a target regulatory element nucleotide sequence of the plant cell such that the target regulatory element nucleotide sequence aligns with the selected variant nucleotide sequence;
- regenerating a genetically modified first plant from the plant cell;
- and
- crossing the genetically modified first plant with a second plant to produce a population of genetically modified plants.
3 more elements
-45. (canceled)
A system for predicting expression of genetic variants, the system comprising a computer-readable medium comprising an artificial intelligence (AI) model, wherein the AI is configured to:
- calculate a fitness score for each of a plurality of variant nucleotide sequences of a plant regulatory element, wherein calculating the fitness score comprises providing the AI model with a first dataset comprising a reference nucleotide sequence of the plant regulatory element and a second dataset comprising the plurality of variant nucleotide sequences of the plant regulatory element and predicting one or more expression profiles of each of the plurality of variant nucleotide sequences relative to expression of the reference nucleotide sequence;
- and
- select a variant nucleotide sequence from the plurality of variant nucleotide sequences based on the fitness score.
-100. (canceled)
Dependencies are inferred from the claim text.