# US 11,908,547 B2 · Methods and compositions for governing phenotypic outcomes in plants > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,908,547 B2 - Document ID: US11908547B2 - Status: Granted - Kind: B2 - Application number: 16/870,838 - Filed: 2020-05-08 - Published: 2020-11-12 - Granted: 2024-02-20 - Earliest claimed: 2019-05-08 - Assignees: X Development LLC - Applicants: X Development LLC - Inventors: Bradley Michael Zamft; Logan Graham - Web page: https://patentgrep.com/patent/US11908547B2 ## CPC - G16B 20/00: ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations - G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] - G06N 3/0475: Generative networks - G06N 3/09: Supervised learning - G06N 3/094: Adversarial learning - G06N 5/04: Inference or reasoning models - G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks - G06N 20/00: Machine learning - G16B 20/50: Mutagenesis - G16B 40/00: ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding - G06N 3/045: Combinations of networks - G06N 3/047: Probabilistic or stochastic networks - G06Q 50/02: Agriculture; Fishing; Forestry; Mining - G16B: BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR ## Abstract Methods, systems, and apparatus, including computer programs encoded on computer storage media, for governing phenotypic outcomes in plants. One method includes obtaining a model input comprising time series data, wherein the time series data comprises, for each previous time point of one or more previous time points, at least one of i) first multi-omics data corresponding to a plant at the previous time point, or ii) phenotypic data corresponding to the plant at the previous time point; and processing the model input using a machine learning model to obtain a model output that comprises, for each future time point of one or more of future time points, a prediction of at least one of i) a phenotype of the plant at the future time point, or ii) second multi-omics data corresponding to the plant at the future time point. ## Claims ### Claim 1 A method performed by one or more computers, the method comprising: obtaining a model input comprising time series data characterizing a plant over a sequence of multiple previous time points, wherein the time series data comprises, for each previous time point in the sequence of multiple previous time points, at least one of: i) multi-omics data corresponding to the plant at the previous time point, or ii) phenotypic data corresponding to the plant at the previous time point; and processing the model input using a machine learning model to generate a model output that comprises, for each future time point of one or more of future time points, a prediction of at least one of: i) predicted phenotypic data corresponding to the plant at the future time point, or ii) predicted multi-omics data corresponding to the plant at the future time point; wherein the machine learning model has been configured through training by a machine learning training technique to: receive a training model input corresponding to a training plant, and process the training model input to generate a prediction corresponding to the training plant at each of one or more future time points; and processing the model output, generated by the machine learning model based on the time series data characterizing the plant over the sequence of multiple time points, to select one or more interventions to be applied to the plant. ### Claim 2 (depends on 1) The method of claim 1, wherein processing the model output, generated by the machine learning model based on the time series data characterizing the plant over the sequence of multiple time points, to select one or more interventions to be applied to the plant comprises selecting an intervention comprising one or more of: modifying the genome of the plant; editing the RNA of the plant; modifying the levels of particular transcripts in the plant; or delivering proteins to the plant. ### Claim 3 (depends on 1) The method of claim 1, wherein processing the model output, generated by the machine learning model based on the time series data characterizing the plant over the sequence of multiple time points, to select one or more interventions to be applied to the plant comprises one or more of: determining a management practice profile of the plant using the model output; determining an environment profile of the plant using the model output; or determining a breeding decision for the plant using the model output. ### Claim 4 (depends on 1) The method of claim 1, wherein the time series data further comprises, for each previous time point in the sequence of multiple previous time points, one or more of environment or management practice data corresponding to the previous time point. ### Claim 5 (depends on 1) The method of claim 1, wherein for each of one or more previous time points: the model input comprises multi-omics data corresponding to the plant at the previous time point; and the multi-omics data corresponding to the plant at the previous time point comprises one or more of: genome, epigenome, transcriptome, proteome, metabolome, lipidome, glycome, cytome, exome, interferome, kinome, ionome, metalome, methylome, phenome, phytochemome, regulome, secretome data, meta-genome, meta-epigenome, meta-transcriptome, meta-proteome, meta-metabolome, meta-lipidome, meta-glycome, meta-cytome, meta-exome, meta-interferome, meta-kinome, meta-ionome, meta-metalome, meta-methylome, meta-phenome, meta-phytochemome, meta-regulome, or meta-secretome data. ### Claim 6 (depends on 1) The method of claim 1, wherein the machine learning model is a hybrid model, wherein the hybrid model comprises i) a machine-learned component and ii) an encoded component. ### Claim 7 One or more non-transitory computer storage media encoded with computer program instructions that when executed by a plurality of computers cause the plurality of computers to perform operations comprising: obtaining a model input comprising time series data characterizing a plant over a sequence of multiple previous time points, wherein the time series data comprises, for each previous time point in the sequence of multiple previous time points, at least one of: i) multi-omics data corresponding to the plant at the previous time point, or ii) phenotypic data corresponding to the plant at the previous time point; and processing the model input using a machine learning model to generate a model output that comprises, for each future time point of one or more of future time points, a prediction of at least one of: i) predicted phenotypic data corresponding to the plant at the future time point, or ii) predicted multi-omics data corresponding to the plant at the future time point; wherein the machine learning model has been configured through training by a machine learning training technique to: receive a training model input corresponding to a training plant, and process the training model input to generate a prediction corresponding to the training plant at each of one or more future time points; and processing the model output, generated by the machine learning model based on the time series data characterizing the plant over the sequence of multiple time points, to select one or more interventions to be applied to the plant. ### Claim 8 (depends on 7) The non-transitory computer storage media of claim 7, wherein processing the model output, generated by the machine learning model based on the time series data characterizing the plant over the sequence of multiple time points, to select one or more interventions to be applied to the plant comprises selecting an intervention comprising one or more of: modifying the genome of the plant; editing the RNA of the plant; modifying the levels of particular transcripts in the plant; or delivering proteins to the plant. ### Claim 9 (depends on 7) The non-transitory computer storage media of claim 7, wherein processing the model output, generated by the machine learning model based on the time series data characterizing the plant over the sequence of multiple time points, to select one or more interventions to be applied to the plant comprises one or more of: determining a management practice profile of the plant using the model output; determining an environment profile of the plant using the model output; or determining a breeding decision for the plant using the model output. ### Claim 10 (depends on 7) The non-transitory computer storage media of claim 7, wherein the time series data further comprises, for each previous time point in the sequence of multiple previous time points, one or more of environment or management practice data corresponding to the previous time point. ### Claim 11 (depends on 7) The non-transitory computer storage media of claim 7, wherein for each of one or more previous time points: the model input comprises multi-omics data corresponding to the plant at the previous time point; and the multi-omics data corresponding to the plant at the previous time point comprises one or more of: genome, epigenome, transcriptome, proteome, metabolome, lipidome, glycome, cytome, exome, interferome, kinome, ionome, metalome, methylome, phenome, phytochemome, regulome, secretome data, meta-genome, meta-epigenome, meta-transcriptome, meta-proteome, meta-metabolome, meta-lipidome, meta-glycome, meta-cytome, meta-exome, meta-interferome, meta-kinome, meta-ionome, meta-metalome, meta-methylome, meta-phenome, meta-phytochemome, meta-regulome, or meta-secretome data. ### Claim 12 (depends on 7) The non-transitory computer storage media of claim 7, wherein the machine learning model is a hybrid model, wherein the hybrid model comprises i) a machine-learned component and ii) an encoded component. ### Claim 13 A system comprising: one or more computers; and one or more storage devices communicatively coupled to the one or more computers, wherein the one or more storage devices store instructions that, when executed by the one or more computers, cause the one or more computers to perform operations comprising: obtaining a model input comprising time series data characterizing a plant over a sequence of multiple previous time points, wherein the time series data comprises, for each previous time point in the sequence of multiple previous time points, at least one of: i) multi-omics data corresponding to the plant at the previous time point, or ii) phenotypic data corresponding to the plant at the previous time point; and processing the model input using a machine learning model to generate a model output that comprises, for each future time point of one or more of future time points, a prediction of at least one of: i) predicted phenotypic data corresponding to the plant at the future time point, or ii) predicted multi-omics data corresponding to the plant at the future time point; wherein the machine learning model has been configured through training by a machine learning training technique to: receive a training model input corresponding to a training plant, and process the training model input to generate a prediction corresponding to the training plant at each of one or more future time points; and processing the model output, generated by the machine learning model based on the time series data characterizing the plant over the sequence of multiple time points, to select one or more interventions to be applied to the plant. ### Claim 14 (depends on 13) The system of claim 13, wherein processing the model output, generated by the machine learning model based on the time series data characterizing the plant over the sequence of multiple time points, to select one or more interventions to be applied to the plant comprises selecting an intervention comprising one or more of: modifying the genome of the plant; editing the RNA of the plant; modifying the levels of particular transcripts in the plant; or delivering proteins to the plant. ### Claim 15 (depends on 13) The system of claim 13, wherein processing the model output, generated by the machine learning model based on the time series data characterizing the plant over the sequence of multiple time points, to select one or more interventions to be applied to the plant comprises one or more of: determining a management practice profile of the plant using the model output; determining an environment profile of the plant using the model output; or determining a breeding decision for the plant using the model output. ### Claim 16 (depends on 13) The system of claim 13, wherein the time series data further comprises, for each previous time point in the sequence of multiple previous time points, one or more of environment or management practice data corresponding to the previous time point. ### Claim 17 (depends on 13) The system of claim 13, wherein for each of one or more previous time points: the model input comprises multi-omics data corresponding to the plant at the previous time point; and the multi-omics data corresponding to the plant at the previous time point comprises one or more of: genome, epigenome, transcriptome, proteome, metabolome, lipidome, glycome, cytome, exome, interferome, kinome, ionome, metalome, methylome, phenome, phytochemome, regulome, secretome data, meta-genome, meta-epigenome, meta-transcriptome, meta-proteome, meta-metabolome, meta-lipidome, meta-glycome, meta-cytome, meta-exome, meta-interferome, meta-kinome, meta-ionome, meta-metalome, meta-methylome, meta-phenome, meta-phytochemome, meta-regulome, or meta-secretome data. ### Claim 18 (depends on 13) The system of claim 13, wherein the machine learning model is a hybrid model, wherein the hybrid model comprises i) a machine-learned component and ii) an encoded component. ### Claim 19 (depends on 13) The system of claim 13, wherein the machine learning model comprises an artificial neural network. ### Claim 20 (depends on 13) The system of claim 13, wherein for each previous time point in the sequence of multiple previous time points, the time series data comprises phenotypic data corresponding to the plant at the previous time point.