Alteration of plant genomes based on assessment of RNA expression outliers

Abstract

The present disclosure provides methods of improving performance of a plant by assessing outlier status of RNA transcripts within the plant or plant population, determining genetic modifications to reduce outlier status of selected loci, and introducing into a plant the determined genetic modifications. In addition, methods are provided of identifying targets for genetic modification to improve performance based on assessment of outlier status of RNA transcripts. Also provided herein are methods of selected two plants to breed or hybridize based on assessment of outlier status of RNA transcripts. A plant, seed, or plant grown from a seed produced by the provided methods are also included herein.

Description

ALTERATION OF PLANT GENOMES BASED ON ASSESSMENT OF RNA EXPRESSION OUTLIERSCROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Patent Application No. 63 / 472,545, filed on June 12, 2023, the entirety of which is incorporated herein by reference.FIELD

[0002] The present disclosure relates generally to methods of improving plant performance, and more specifically to methods of improving plant performance based on RNA expression outlier loci.BACKGROUND

[0003] Traditional phenotype-based breeding and the advent of genomic technologies have resulted in significant advances in improving the performance of economically valuable crops. Whereas some beneficial traits may be controlled by single loci, such as certain disease resistance genes, other traits such as yield performance in maize are highly polygenic, or controlled by more than 50 genomic loci (Heffner et al. (2009) Crop Science 49(1): 1-12). Stably and reliably improving performance of highly polygenic traits is often more complicated than performing simple crosses, or making a single genomic modification.

[0004] In the process of producing domestication and stable crop lines, many loci in modem day crops have been fixed for particular alleles. Some alleles that are fixed in the population may affect performance in non-optimal ways (Kremling et al. (2018) Nature 555(7697): 520-523, and Zhao et al. (2016) The American Journal of Human Genetics 98(2): 299-309).

[0005] With recent improvements to the field of genomic modifications and editing, it is now possible to make numerous concurrent genomic modifications. However, better methods are needed for selecting a multitude of genomic targets in order to improve plant performance.BRIEF SUMMARY

[0006] In one aspect, the present disclosure provides a method for improving performance of a plant, comprising: a. providing RNA expression data for a plurality of RNAtranscripts of the plant or of a plant population comprising the plant; b. determining an outlier score for the expression level of each RNA transcript of the plurality of RNA transcripts in comparison to one or more reference transcript expression levels; c. obtaining a distribution of the outlier scores; d. identifying a subset of RNA transcripts with outlier scores beyond a threshold in the distribution; e. determining for each RNA transcript of the subset of RNA transcripts a genetic modification that changes the RNA expression of the RNA transcript to have a reduced outlier score magnitude; and f. introducing into the plant the genetic modifications determined for each RNA transcript of the subset of RNA transcripts, thereby improving the performance of the plant. In one aspect, the present disclosure provides a method for improving performance of a plant, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of the plant or of a plant population comprising the plant; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of RNA transcript in the plant or plant population in comparison to one or more reference transcript expression levels; c) obtaining a distribution of the plurality of outlier scores; d) identifying a subset of the plurality of RNA transcripts with the plurality of outlier scores beyond a threshold in the distribution; e) determining, for each RNA transcript of the subset of RNA transcripts, a genetic modification that changes the RNA expression of the RNA transcript to have a reduced outlier score magnitude; and f) introducing into the plant the genetic modifications determined for one or more of the RNA transcripts of the subset of RNA transcripts, thereby improving the performance of the plant. In some embodiments, the method further comprises the step of: g) comparing the performance of the plant comprising the determined genetic modifications to the performance of a control plant. In some embodiments, the control plant is an unedited plant from the same population, grown under the same conditions. In some embodiments, the control plant is a plant from a different cultivar, grown under the same conditions. In some embodiments, step f) comprises introducing into the plant the genetic modifications determined for at least 10%, 20%, 30%, 40%, 50%, 60%, 70%, 75%, 80%, 85%, 90%, 95%, or 100% of the transcripts of the subset of RNA transcripts. In some embodiments, step f) comprises introducing into the plant the genetic modifications determined for each transcript of the subset of RNA transcripts.

[0007] In another aspect, the present disclosure provides a method of identifying targets for genetic modification to improve plant performance, comprising: a. providing RNA expression data for a plurality of RNA transcripts of the plant or of a plant population comprising the plant; b. determining an outlier score for the expression level of each RNAtranscript of the plurality of RNA transcripts in comparison to one or more reference transcript expression levels; c. obtaining a distribution of the outlier scores; d. identifying a subset of RNA transcripts with outlier scores beyond a threshold in the distribution; and e. determining for each RNA transcript of the subset of RNA transcripts a genetic modification that changes the RNA expression of the RNA transcript to have a reduced outlier score magnitude, thereby identifying targets for genetic modification to improve plant performance. In another aspect, the present disclosure provides a method of identifying targets for genetic modification to improve performance of a plant, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of the plant or of a plant population comprising the plant; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in the plant or plant population in comparison to one or more reference transcript expression levels; c) obtaining a distribution of the plurality of outlier scores; d) identifying a subset of the plurality of RNA transcripts with outlier scores beyond a threshold in the distribution; and e) determining, for each RNA transcript of the subset of RNA transcripts, a genetic modification that changes the RNA expression of the RNA transcript to have a reduced outlier score magnitude, thereby identifying targets for genetic modification to improve plant performance. In some embodiments, the method further comprises the step of: f) selecting a subset of targets for genetic modification in a plant from the determined genetic modifications, wherein the selected genetic modifications are predicted to reduce the magnitude of RNA expression outlier scores of the RNA transcripts.

[0008] In another aspect, the present disclosure provides a method for selecting two plant individuals to breed or hybridize, comprising: a. providing RNA expression data for a plurality of RNA transcripts of more than one plant or plant population; b. determining an outlier score for the expression level of each RNA transcript of the plurality of RNA transcripts in comparison to one or more reference transcript expression levels for each of the more than one plants or plant populations; c. obtaining a distribution of the outlier scores for each of the more than one plants or plant populations; d. identifying in each of the more than one plants or plant populations a subset of RNA transcripts with outlier scores beyond a specified threshold in the distribution; and e. selecting two plant individuals capable of breeding or hybridizing with one another, wherein the two plant individuals have minimal overlap in their two subsets of RNA transcripts, or wherein the two plant individuals are derived from two plant populations with minimal overlap in their two subsets of RNAtranscripts. In another aspect, the present disclosure provides a method for selecting two plant individuals to breed or hybridize, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of more than one plant or plant population; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in each of the more than one plants or plant populations in comparison to one or more reference transcript expression levels for each of the more than one plants or plant populations; c) obtaining a distribution of the plurality of outlier scores for each of the more than one plants or plant populations; d) identifying in each of the more than one plants or plant populations a subset of the plurality of RNA transcripts with outlier scores beyond a specified threshold in the distribution; and e) selecting two plant individuals capable of breeding or hybridizing with one another, wherein the two plant individuals have minimal overlap in their two subsets of RNA transcripts, or wherein the two plant individuals are derived from two plant populations with minimal overlap in their two subsets of RNA transcripts. In some embodiments, one or more RNA transcripts present in the subset of RNA transcripts of the first of the two selected plant individuals are not present in the subset of RNA transcripts of the second of the two selected plant individuals. In some embodiments, between two and twenty RNA transcripts present in the subset of RNA transcripts of the first of the two selected plant individuals are not present in the subset of RNA transcripts of the second of the two selected plant individuals. In some embodiments, the selected two plants capable of breeding or hybridizing with one another have less than 50% of the subset of RNA transcripts overlapping between their two subsets of RNA transcripts. In some embodiments, the selected two plants capable of breeding or hybridizing with one another have no overlap in their two subsets of RNA transcripts.

[0009] In some embodiments of any of the previous aspects, the method further comprises steps e’), e”), and e”’) after step e) and before any following steps, comprising: e’) analyzing the effect of one or more genetic modifications determined in step e) on RNA expression; e”) determining one or more alternative genetic modifications for an RNA transcript and performing additional analysis; and e”’) selecting a genetic modification for each RNA transcript in the subset of RNA transcripts. In some embodiments, the analysis is performed in vitro. In some embodiments, the analysis is performed in silico. In some embodiments, the analysis is performed in cell culture. In some embodiments, the analysis is performed in protoplasts. In some embodiments, the analysis is performed by Perturb-Seq. In some embodiments, step e”) is repeated until a genetic modification with a desired outcomeis obtained. In some embodiments, the desired outcome is a reduction of the numeric outlier score for the RNA transcript such that the outlier score is no longer beyond the designated threshold. In some embodiments, the desired outcome is a reduction of the numeric outlier score for the RNA transcript such that the outlier score is no longer beyond the threshold.

[0010] In some embodiments of any of the previous aspects, the RNA expression data is not pooled from multiple individuals. In some embodiments of any of the previous aspects, the RNA expression data is combined from multiple individuals from the plant population. In some embodiments of any of the previous aspects, the outlier score is determined using RNA expression data across multiple tissues. In some embodiments, the RNA expression data are from a sample comprising multiple tissues. In some embodiments, the RNA expression data is pooled from multiple tissues. In some embodiments of any of the previous aspects, the outlier score is determined using RNA expression data from a single tissue. In some embodiments of any of the previous aspects, the outlier score is determined using RNA expression data across multiple members of a gene interaction network. In some embodiments of any of the previous aspects, the outlier score is determined using protein expression data in addition to RNA expression data. In some embodiments of any of the previous aspects, the outlier score is determined using splice isoform ratios. In some embodiments of any of the previous aspects, the RNA expression data is whole genome RNA-seq data. In some embodiments of any of the previous aspects, the RNA expression data is derived from micro level tissue. In some embodiments, the micro level tissue is selected from the group consisting of shoot apical meristem, inflorescence meristem, immature male embryos, immature female embryos, developing anthers, and root hairs. In some embodiments, the RNA expression data is obtained by single cell sequencing. In some embodiments, the micro level tissue is obtained by laser capture microdissection. In some embodiments of any of the previous aspects, the RNA expression data is derived from tissue isolated at a specific plastochron. In some embodiments of any of the previous aspects, the RNA expression data is derived from macro level tissue. In some embodiments, the macro level tissue is selected from the group consisting of leaf, root, internode, anther, tassel, ear shank, kernel, and ligule. In some embodiments of any of the previous aspects, the RNA expression data is derived from protoplasts. In some embodiments of any of the previous aspects, the RNA expression data is derived from undifferentiated cells. In some embodiments, the RNA expression data is derived from plant calli. In some embodiments, the RNA expression data is derived from plant cell suspension cultures. In some embodiments of any of the previous aspects, the RNA expression levels are not pooled from multipleindividuals. In some embodiments of any of the previous aspects, the RNA expression levels are combined from multiple individuals from the plant population. In some embodiments of any of the previous aspects, the outlier score is determined using RNA expression data across multiple tissues. In some embodiments, the RNA expression data are from a sample comprising multiple tissues. In some embodiments, the RNA expression data are pooled from multiple tissues. In some embodiments of any of the previous aspects, the outlier score is determined using RNA expression levels from a single tissue. In some embodiments of any of the previous aspects, the outlier score is determined using RNA expression levels across multiple members of a gene interaction network. In some embodiments of any of the previous aspects, the outlier score is determined using protein expression data in addition to RNA expression levels. In some embodiments of any of the previous aspects, the outlier score is determined using splice isoform ratios. In some embodiments of any of the previous aspects, the RNA expression levels are whole genome RNA-seq data. In some embodiments of any of the previous aspects, the RNA expression levels are derived from micro level tissue. In some embodiments, the micro level tissue is selected from the group consisting of shoot apical meristem, inflorescence meristem, immature male embryos, immature female embryos, developing anthers, and root hairs. In some embodiments, the RNA expression levels are obtained by single cell sequencing. In some embodiments, the micro level tissue is obtained by laser capture microdissection. In some embodiments of any of the previous aspects, the RNA expression levels are derived from tissue isolated at a specific plastochron. In some embodiments of any of the previous aspects, the RNA expression levels are derived from macro level tissue. In some embodiments, the macro level tissue is selected from the group consisting of leaf, root, internode, anther, tassel, ear shank, kernel, and ligule. In some embodiments of any of the previous aspects, the RNA expression levels are derived from protoplasts. In some embodiments of any of the previous aspects, the RNA expression levels are derived from undifferentiated cells. In some embodiments, the RNA expression levels are derived from plant calli. In some embodiments, the RNA expression levels are derived from plant cell suspension cultures. In some embodiments of any of the previous aspects, the plurality of RNA transcripts comprises mRNAs. In some embodiments of any of the previous aspects, the plurality of RNA transcripts comprises IncRNAs. In some embodiments of any of the previous aspects, the plurality of RNA transcripts comprises miRNAs. In some embodiments of any of the previous aspects, the plurality of RNA transcripts comprises siRNAs. In some embodiments of any of the previous aspects, the plurality of RNA transcripts comprises transcripts associated with flowering. In some embodiments of any ofthe previous aspects, the plurality of RNA transcripts comprises transcripts associated with growth. In some embodiments of any of the previous aspects, the plurality of RNA transcripts comprises transcripts associated with biotic stress resistance. In some embodiments of any of the previous aspects, the plurality of RNA transcripts comprises transcripts associated with abiotic stress resistance. In some embodiments of any of the previous aspects, the plurality of RNA transcripts consists of all RNA transcripts.

[0011] In some embodiments of any of the previous aspects, the reference transcript expression level is the transcript level from another plant of the same species. In some embodiments of any of the previous aspects, the reference transcript expression level is the RNA expression level of the RNA transcript in another plant of the same species. In some embodiments of any of the previous aspects, the reference transcript expression level is the consensus transcript level within a population of plants of the same species. In some embodiments of any of the previous aspects, the reference transcript expression level is the consensus transcript level across multiple cultivars of the same species. In some embodiments of any of the previous aspects, the reference transcript expression level is the consensus transcript level of orthologs across multiple species. In some embodiments, the consensus transcript level is the mean level. In other embodiments, the consensus transcript is the median level.

[0012] In some embodiments of any of the previous aspects, the outlier score is expressed as the deviation of the mean scaled by the standard deviation of the population. In some embodiments, the outlier score is a Z-score. In some embodiments, the outlier score is a burden score tied to a Z-score. In some embodiments, the Z-score is a non-parametric Z- score. In some embodiments, the Z-score is derived from provided estimates of the dispersal or from the defined point mass of the distribution.

[0013] In some embodiments of any of the previous aspects, the outlier score is determined by autoencoder-based outlier detection, wherein the autoencoder-based outlier detection comprises an Encoder and a Decoder, wherein the Encoder and the Decoder each comprise non-linear transformations. In some embodiments, the Encoder and the Decoder each comprise one or more non-linear transformations. In some embodiments, the autoencoder takes an input. In some embodiments, the autoencoder takes an input sample. In some embodiments, the input to the autoencoder comprises RNA expression measurements. In some embodiments, the Encoder of the autoencoder is an artificial neural network (ANN). In some embodiments, the Encoder of the autoencoder comprises an artificial neural network (ANN). In some embodiments, the Decoder of the autoencoder is an artificial neural network(ANN). In some embodiments, the Decoder of the autoencoder comprises an artificial neural network (ANN). In some embodiments, the non-linear transformations of the Encoder, the Decoder, or both the Encoder and Decoder are modified by tuning a set of parameters (9). In some embodiments, the parameters (9) are trained by passing an input vector through the model to produce an output vector, and iteratively tuning the parameters (9) via backpropagation through the series of non-linear transformations by minimizing a differentiable reconstruction error function between the output vector and the input vector. In some embodiments, the differentiable reconstruction error function is the mean squared error (MSE). In some embodiments, the outlier score is determined by anomaly detection with attributed networks. In some embodiments, the outlier score is determined for a set of genes in a known gene regulatory network. In some embodiments, the input to the autoencoder comprises a vector comprising measured expression values for a single gene across multiple tissues. In some embodiments, the input to the autoencoder comprises a vector comprising measured expression values for a single gene across multiple tissues. In some embodiments, the Encoder is constructed as a graph neural network (GNN). In some embodiments, the Decoder is constructed as a graph neural network (GNN). In some embodiments, the nonlinear transformations of the Encoder, the Decoder, or both the Encoder and Decoder are modified by tuning a set of parameters (9). In some embodiments, the parameters (9) are trained by passing an input vector through the model to produce an output vector, and iteratively tuning the parameters (9) via backpropagation through the series of non-linear transformations by minimizing a differentiable reconstruction error function between the output vector and the input vector. In some embodiments, the differentiable reconstruction error function is the mean squared error (MSE).

[0014] In some embodiments of any of the previous aspects, the outlier score is determined by multivariate outlier detection methods, wherein information is borrowed across genes in a local network and across tissues. In some embodiments, the multivariate outlier detection method comprises use of the Mahalanobis distance.

[0015] In some embodiments of any of the previous aspects, the threshold is a one-tailed threshold on the distribution. In some embodiments of any of the previous aspects, the threshold is a two-tailed threshold on the distribution. In some embodiments of any of the previous aspects, the threshold is a p-value. In some embodiments of any of the previous aspects, the threshold is a set number of genes at one or more tails of the distribution. In some embodiments, the threshold is set at 50 genes at the upper tail of the distribution.

[0016] In some embodiments of any of the previous aspects, the genetic modification determined for an individual RNA transcript creates a novel allele. In some embodiments of any of the previous aspects, the genetic modification determined for an individual RNA transcript replicates an alternative allele not present in the plant. In some embodiments, the alternative allele is an allele present in the plant from which the reference transcript expression level is derived. In some embodiments, the alternative allele is an allele with a reduced outlier score from another plant.

[0017] In some embodiments of any of the previous aspects, the genetic modification determined for an individual RNA transcript targets a DNA sequence element selected from the group consisting of a promoter, an enhancer, a transcription factor binding site, a splice donor site, a splice acceptor site, an miRNA binding site, an siRNA binding site, an insulator, an intron, an exon, a 5’ UTR, and a 3’ UTR. In some embodiments, the genetic modification determined for an individual RNA transcript is a small deletion in the promoter. In some embodiments, the genetic modification determined for an individual RNA transcript is a small element insertion to increase expression.

[0018] In some embodiments of any of the previous aspects, the genome of the plant is modified using a nuclease selected from the group consisting of a Cas nuclease, a TALE nuclease, a zinc finger nuclease, a meganuclease, or an engineered nuclease. In some embodiments, the Cas nuclease is a Cas9 or Cas 12 nuclease.

[0019] In some embodiments of any of the previous aspects, the reduced outlier score magnitude of an RNA transcript after genetic modification is no longer beyond the determined threshold. In some embodiments of any of the previous aspects, the reduced outlier score magnitude of an RNA transcript after genetic modification is no longer beyond the threshold. In some embodiments of any of the previous aspects, the reduced outlier score magnitude is reduced by at least one standard deviation of the distribution.

[0020] In some embodiments of any of the previous aspects, the plant is selected from the group consisting of corn, soy, and wheat.

[0021] In some embodiments of any of the previous aspects, the RNA expression data is obtained from a plant from an inbred panel or a hybrid panel. In some embodiments of any of the previous aspects, the RNA expression data is obtained from a plant from an asexually- propagated population. In some embodiments of any of the previous aspects, the RNA expression levels are obtained from a plant from an inbred panel or a hybrid panel. In some embodiments of any of the previous aspects, the RNA expression levels are obtained from a plant from an asexually-propagated population.

[0022] In some embodiments of any of the previous aspects, improved performance of the plant is measured by a metric selected from the group consisting of growth, yield, survival, biomass, number of kernels, ear weight, and nutritional content.

[0023] In another aspect, the present disclosure provides a plant produced by the method of any of the previous aspects.

[0024] In another aspect, the present disclosure provides a seed produced by the method of nay of the previous aspects. In a further aspect, the present disclosure provides a plant grown by the seed produced by the method of any of the previous aspects.DETAILED DESCRIPTION

[0025] The following description sets forth exemplary methods, parameters and the like. It should be recognized, however, that such description is not intended as a limitation on the scope of the present disclosure but is instead provided as a description of exemplary embodiments.Definitions

[0026] As used herein, the terms “a,” “an,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. Where a term is provided in the singular, the inventors also contemplate aspects of the invention described by the plural of that term.

[0027] The phrase “allelic variant” as used herein refers to a polynucleotide or polypeptide sequence variant that occurs in a different strain, variety, or isolate of a given organism.

[0028] The term "and / or" where used herein is to be taken as specific disclosure of each of the two specified features or components with or without the other. Thus, the term “and / or" as used in a phrase such as "A and / or B" herein is intended to include "A and B," "A or B," "A" (alone), and "B" (alone). Likewise, the term "and / or" as used in a phrase such as "A, B, and / or C" is intended to encompass each of the following embodiments: A, B, and C; A, B, or C; A or C; A or B; B or C; A and C; A and B; B and C; A (alone); B (alone); and C (alone).

[0029] As used herein, the terms “comprise,” comprises, “comprising,” “include,” “includes,” and “including” can be interchanged and are to be construed as at least having the features to which they refer while not excluding any additional unspecified features.

[0030] As used herein, the terms “crossing,” “breeding,” and “hybridizing” can be interchanged, and refer to the process of two plant performing sexual reproduction to produce offspring. When used as a verb, these terms assume human guidance or assistance in performing sexual reproduction of the chosen plants.

[0031] Although the following description uses terms “first,” “second,” etc. to describe various elements, these elements should not be limited by the terms. These terms are only used to distinguish one element from another. For example, a first plant could be termed a second plant, and, similarly, a second plant could be termed a first plant, without departing from the scope of the various described embodiments. The first plant and the second plant are both plants, but they are not the same plant.

[0032] As used herein, the term “hybrid panel” refers to a selection of one or more plants or plant genotypes generated by crossing two inbred lines. Plants or plant genotypes of a hybrid panel are heterozygous at the genomic locations where the two inbred parents carry different alleles.

[0033] As used herein, the term “inbred panel” refers to a selection of one or more plants or plant genotypes that are developed for use as parents in the production of hybrid cultivars and synthetic cultivars in the breeding of cross-pollinated species.

[0034] As used herein, the terms “include,” “includes,” and “including” are to be construed as at least having the features to which they refer while not excluding any additional unspecified features.

[0035] As used herein, the terms “orthologous,” “ortholog,” and “orthologue” are used to describe genes or the RNAs or proteins encoded by those genes that are from different species but which have the same function (e.g., encode RNAs which exhibit the same meristem transport function). Orthologous genes will typically encode RNAs or proteins with some degree of sequence identity and can also exhibit conservation of sequence motifs, and / or conservation of structural features including RNA stem loop structures.

[0036] As used herein, the term “plant” includes a whole plant and any descendant, cell, tissue, or part of a plant. The term “plant parts” include any part(s) of a plant, including, for example and without limitation: seed (including mature seed and immature seed); a plant cutting; a plant cell; a plant cell culture; or a plant organ (e.g., intact nodal bud, shoot apex or shoot apical meristem, root apex or root apical meristem, lateral meristem, intercalary meristem, zygotic embryo, somatic embryo, ovule, pollen, microspore, anther, tassel, ear shank, kernel, hypocotyl, cotyledon, leaf, petiole, stem, internode, tuber, root, flowers, fruits,shoots, and explants). A plant tissue or plant organ may be a seed, protoplast, callus, or any other group of plant cells that is organized into a structural or functional unit. A plant cell or tissue culture may be capable of regenerating a plant having the physiological and morphological characteristics of the plant from which the cell or tissue was obtained, and of regenerating a plant having substantially the same genotype as the plant. Regenerable cells in a plant cell or tissue culture may be embryos, protoplasts, meristematic cells, callus, pollen, leaves, anthers, roots, root tips, silk, flowers, kernels, ears, cobs, husks, or stalks. In contrast, some plant cells are not capable of being regenerated to produce plants and are referred to herein as “non-regenerable” plant cells.

[0037] As used herein, the term “plastochron” refers to a unit of time corresponding to the interval between two successive similar, periodically repeated events. For example, the time between the appearance of one leaf primordia and the appearance of the next leaf primordia is a plastochron.

[0038] As used herein, the phrase “transcript” refers to a ribonucleic acid polynucleotide produced by the process of transcription from the genomic DNA of an organism. Multiple different transcripts may be produced from the same genomic locus, as in the non-limiting example of a gene that undergoes alternative splicing of introns and exons, thus producing multiple different mRNA transcripts. Products of alternative splicing may also be referred to as “isoforms”. Isoforms may be analyzed together or individually, depending on the purpose of the analysis. In another non-limiting example, a locus may produce both a messenger mRNA and a regulatory long non-coding IncRNA, both of which are transcripts of the locus.

[0039] As used herein, the phrase “sequence identity” refers to the percent similarity of two polynucleotides or polypeptides. A polynucleotide or polypeptide has a certain percent “sequence identity” to another polynucleotide or polypeptide, meaning that, when aligned, that percentage of bases or amino acids are the same, and in the same relative position, when comparing the two sequences. Sequence similarity can be determined in a number of different manners. To determine sequence identity, sequences can be aligned using the methods and computer programs, including BLAST, available at ncbi[dot]nlm[dot]nih[dot]gov / BLAST. See, e.g., Altschul et al. Mol. Biol. 1990, 215:403-410. Another alignment algorithm is FASTA, available in the Genetics Computing Group (GCG) package, from Madison, Wis., USA, a wholly owned subsidiary of Oxford Molecular Group, Inc. Other techniques for alignment are described in Methods in Enzymology, vol. 266: Computer Methods for Macromolecular Sequence Analysis (1996), ed. Doolittle, Academic Press, Inc., a division of Harcourt Brace & Co., San Diego, Calif., USA. Of particular interest are alignment programsthat permit gaps in the sequence. The Smith-Waterman is one type of algorithm that permits gaps in sequence alignments. See Meth. Mol. Biol., 70: 173-187 (1997). Also, the GAP program using the Needleman and Wunsch alignment method can be utilized to align sequences. See Mol. Biol., 48: 443-453 (1970).

[0040] Unless otherwise stated, nucleic acid sequences in the text of this specification are given, when read from left to right, in the 5' to 3' direction. Nucleic acid sequences may be provided as DNA or as RNA, as specified; disclosure of one necessarily defines the other, as well as necessarily defines the exact complements, as is known to one of ordinary skill in the art.

[0041] To the extent to which any of the preceding definitions is inconsistent with definitions provided in any patent or non-patent reference incorporated herein by reference, any patent or non-patent reference cited herein, or in any patent or non-patent reference found elsewhere, it is understood that the preceding definition will be used herein.

[0042] As used herein, a modifier to the initial item in a list is taken to apply to each item in the list. For example, the phrase “at least about 10%, 20%, or 30%” herein is intended to have the same meaning as “at least about 10%, at least about 20%, or at least about 30%. ”Methods

[0043] In one aspect, the present disclosure provides a method for improving performance of a plant, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of the plant or of a plant population comprising the plant; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of RNA transcript in the plant or plant population in comparison to one or more reference transcript expression levels; c) obtaining a distribution of the plurality of outlier scores; d) identifying a subset of the plurality of RNA transcripts with the plurality of outlier scores beyond a threshold in the distribution; e) determining, for each RNA transcript of the subset of RNA transcripts, a genetic modification that changes the RNA expression of the RNA transcript to have a reduced outlier score magnitude; and f) introducing into the plant the genetic modifications determined for one or more of the RNA transcripts of the subset of RNA transcripts, thereby improving the performance of the plant. In some embodiments, the method further comprises the step of: g) comparing the performance of the plant comprising the determined genetic modifications to the performance of a control plant. In some embodiments, the control plant is an unedited plant from the same population, grown underthe same conditions. In some embodiments, the control plant is a plant from a different cultivar, grown under the same conditions. In some embodiments, step f) comprises introducing into the plant the genetic modifications determined for at least 10%, 20%, 30%, 40%, 50%, 60%, 70%, 75%, 80%, 85%, 90%, 95%, or 100% of the transcripts of the subset of RNA transcripts. In some embodiments, step f) comprises introducing into the plant the genetic modifications determined for each transcript of the subset of RNA transcripts.

[0044] In another aspect, the present disclosure provides a method of identifying targets for genetic modification to improve performance of a plant, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of the plant or of a plant population comprising the plant; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in the plant or plant population in comparison to one or more reference transcript expression levels; c) obtaining a distribution of the plurality of outlier scores; d) identifying a subset of the plurality of RNA transcripts with outlier scores beyond a threshold in the distribution; and e) determining, for each RNA transcript of the subset of RNA transcripts, a genetic modification that changes the RNA expression of the RNA transcript to have a reduced outlier score magnitude, thereby identifying targets for genetic modification to improve plant performance. In some embodiments, the method further comprises the step of: f) selecting a subset of targets for genetic modification in a plant from the determined genetic modifications, wherein the selected genetic modifications are predicted to reduce the magnitude of RNA expression outlier scores of the RNA transcripts.

[0045] In another aspect, the present disclosure provides a method for selecting two plant individuals to breed or hybridize, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of more than one plant or plant population; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in each of the more than one plants or plant populations in comparison to one or more reference transcript expression levels for each of the more than one plants or plant populations; c) obtaining a distribution of the plurality of outlier scores for each of the more than one plants or plant populations; d) identifying in each of the more than one plants or plant populations a subset of the plurality of RNA transcripts with outlier scores beyond a specified threshold in the distribution; and e) selecting two plant individuals capable of breeding or hybridizing with one another, wherein the two plant individuals have minimaloverlap in their two subsets of RNA transcripts, or wherein the two plant individuals are derived from two plant populations with minimal overlap in their two subsets of RNA transcripts. In some embodiments, one or more RNA transcripts present in the subset of RNA transcripts of the first of the two selected plant individuals are not present in the subset of RNA transcripts of the second of the two selected plant individuals. In some embodiments, between two and twenty RNA transcripts present in the subset of RNA transcripts of the first of the two selected plant individuals are not present in the subset of RNA transcripts of the second of the two selected plant individuals. In some embodiments, the selected two plants capable of breeding or hybridizing with one another have less than 50% of the subset of RNA transcripts overlapping between their two subsets of RNA transcripts. In some embodiments, the selected two plants capable of breeding or hybridizing with one another have no overlap in their two subsets of RNA transcripts.

[0046] In some embodiments of any of the previous aspects, the method further comprises steps e’), e”), and e”’) after step e) and before any following steps, comprising: e’) analyzing the effect of one or more genetic modifications determined in step e) on RNA expression; e”) determining one or more alternative genetic modifications for an RNA transcript and performing additional analysis; and e”’) selecting a genetic modification for each RNA transcript in the subset of RNA transcripts. In some embodiments, the analysis is performed in vitro. In some embodiments, the analysis is performed in silico. In some embodiments, the analysis is performed in cell culture. In some embodiments, the analysis is performed in protoplasts. In some embodiments, the analysis is performed by Perturb-Seq. In some embodiments, step e”) is repeated until a genetic modification with a desired outcome is obtained. In some embodiments, the desired outcome is a reduction of the numeric outlier score for the RNA transcript such that the outlier score is no longer beyond the threshold.RNA Expression Data

[0047] In some embodiments, the provided RNA expression data comprises RNA expression levels. In some embodiments, RNA expression levels indicate presence or absence. In some embodiments, RNA expression levels are quantitative measures of RNA. In some embodiments, RNA expression levels are measured in Reads Per Kilobase per Million mapped reads (RPKM).

[0048] In some embodiments, the provided RNA expression data is derived from a single plant. In some embodiments, the provided RNA expression levels are derived from a single plant. In some embodiments, the provided RNA expression data is derived from the plant tobe improved, edited, bred, hybridized, or otherwise selected. In some embodiments, the provided RNA expression levels are derived from the plant to be improved, edited, bred, hybridized, or otherwise selected. In some embodiments, the provided RNA expression data is derived from a plant from the same plant population as the plant to be improved, edited, bred, hybridized, or otherwise selected. In some embodiments, the provided RNA expression levels are derived from a plant from the same plant population as the plant to be improved, edited, bred, hybridized, or otherwise selected. In some embodiments, a plant population is plants of the same species. In some embodiments, a plant population is plants of the same cultivar. In some embodiments, a plant population is plants of the same inbred population. In some embodiments, a plant population is an inbred panel or a hybrid panel. In some embodiments, the genotype of a plant of the inbred panel is homozygous. In some embodiments, the genotype of a plant of the inbred panel is homogenous. In some embodiments, the plant population of the inbred panel is not capable of natural selfing. In some embodiments, the plant population of the inbred panel is artificially maintained by forced selfing. In some embodiments, a plant population is hybrid plants produced by breeding plants from two specific inbred lines. In some embodiments, a plant population is plants grown in the same field. In some embodiments, a plant population is plants from an asexually-propagated population.

[0049] In some embodiments, the provided RNA expression data is combined from multiple individuals from a plant population. In some embodiments, the provided RNA expression levels are combined from multiple individuals from a plant population. In some embodiments, a plant population is plants of the same species. In some embodiments, a plant population is plants of the same cultivar. In some embodiments, a plant population is plants of the same inbred population. In some embodiments, a plant population is plants grown in the same field. In some embodiments, the provided RNA expression data is derived from at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 individual plants. In some embodiments, the provided RNA expression levels are derived from at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or more than 20 individual plants.

[0050] In some embodiments, the RNA expression data is derived from a macro-level tissue. In some embodiments, the RNA expression levels are derived from a macro-level tissue. A “macro-level tissue”, as used herein, refers to a plant tissue or organ composed of a plurality of cell types. In some embodiments, the macro-level tissue can be collected by hand. In some embodiments, the macro-level tissue can be collected by eye, and / or without the useof microscopy. Examples of macro-level tissue include, but are not limited to, leaf, root, internode, anther, tassel, ear shank, kernel, and ligule.

[0051] In some embodiments, the RNA expression data is derived from a micro-level tissue. In some embodiments, the RNA expression levels are derived from a micro-level tissue. A “micro-level tissue”, as used herein, refers to a plant tissue or organ composed of a limited number of cell types. In some embodiments, the micro-level tissue is composed of one, two, three, four, five, six, seven, eight, nine, or ten distinct cell types. In some embodiments, the micro-level tissue is collected or isolated through single cell sorting. In some embodiments, the micro-level tissue is collected or isolated through laser capture microdissection. Examples of micro-level tissue include, but are not limited to, shoot apical meristem, inflorescence meristem, immature male embryos, immature female embryos, developing anthers, and root hairs.

[0052] In some embodiments, the RNA expression data is derived from protoplasts. In some embodiments, the RNA expression levels are derived from protoplasts. A “protoplast”, as used here, refers to a plant cell from which the plant cell wall has been removed. The cell wall can be removed either mechanically, chemically, or enzymatically, using methods known in the art. In some embodiments, the protoplasts are derived from a plant organ. In some embodiments, the protoplasts are derived from an intact nodal bud, shoot apex or shoot apical meristem, root apex or root apical meristem, lateral meristem, intercalary meristem, zygotic embryo, somatic embryo, ovule, pollen, microspore, anther, tassel, ear shank, kernel, hypocotyl, cotyledon, leaf, petiole, stem, internode, tuber, root, flowers, fruits, shoots, or explants.

[0053] In some embodiments, the RNA expression data is derived from undifferentiated cells. In some embodiments, the RNA expression levels are derived from undifferentiated cells. In some embodiments, the undifferentiated cells are plant calli. In some embodiments, the undifferentiated cells are plant cell suspension cultures.

[0054] In some embodiments, the RNA expression data is derived from a tissue collected or isolated at a particular time point. In some embodiments, the RNA expression data is derived from tissue collected or isolated at a specific plastochron. In one embodiment, the RNA expression data is collected or isolated during the plastochron between the appearance of the first leaf primordium and the second leaf primordium. In one embodiment, the RNA expression data is collected or isolated during the plastochron between the appearance of the second leaf primordium and the third leaf primordium. In one embodiment, the RNA expression data is collected or isolated during the plastochron between the appearance of thethird leaf primordium and the fourth leaf primordium. In one embodiment, the RNA expression data is collected or isolated during the plastochron between the appearance of the fourth leaf primordium and the fifth leaf primordium. In some embodiments, the RNA expression levels are derived from a tissue collected or isolated at a particular time point or developmental stage. In some embodiments, the RNA expression levels are derived from tissue collected or isolated at or during a specific plastochron. In one embodiment, the RNA expression levels are collected or isolated during the plastochron between the appearance of the first leaf primordium and the second leaf primordium. In one embodiment, the RNA expression levels are collected or isolated during the plastochron between the appearance of the second leaf primordium and the third leaf primordium. In one embodiment, the RNA expression levels are collected or isolated during the plastochron between the appearance of the third leaf primordium and the fourth leaf primordium. In one embodiment, the RNA expression levels are collected or isolated during the plastochron between the appearance of the fourth leaf primordium and the fifth leaf primordium.

[0055] Methods of obtaining RNA expression data from a plant, plant cell, plant cell culture, or plant tissue sample are well known in the art. RNA can be extracted from fresh or frozen tissue, then sequenced through methods such as RNA-seq. In some embodiments, RNA expression data is obtained by single-cell RNA sequencing. In some embodiments, RNA expression levels are obtained by single-cell RNA sequencing. Methods of analyzing RNA sequencing data, including quality control, trimming, and mapping reads to genomes or known transcriptomes, are known in the art. In some embodiments, the RNA expression data is whole genome RNA sequencing data. In some embodiments, the RNA expression levels are derived from whole genome RNA sequencing data. As used herein, “whole genome RNA sequencing data” encompasses all RNAs present within a cell or tissue sample. RNA extraction, sequencing, and / or analysis methods can be modified to enrich for different types of RNAs, including but not limited to poly adenylated sequences such as messenger RNAs (mRNAs), shorter sequences such as small interfering RNAs (siRNAs) and microRNAs (miRNAs), and nuclear RNAs such as nascent or long non-coding RNAs (IncRNAs).

[0056] Messenger RNAs (mRNAs) are polyadenylated RNA transcripts derived from genes that encode a polypeptide for translation within the cell. mRNAs. mRNAs can be selected for in the RNA sequencing process by means such as, but not limited to, the use of oligodT primers that only bind to poly adenylated transcripts. In some embodiments, the plurality of RNA transcripts comprises mRNAs.

[0057] Long non-coding RNAs (IncRNAs) are RNA transcripts that do not encode a polypeptide. IncRNAs are typically more than 200 nucleotides long, and can be produced from intergenic, intronic, or coding regions of the genome (See Wang and Chekanova (2017) Adv Exp Med Biol 1008: 133-154). IncRNAs can be present in either the nuclear or the cytoplasmic compartments, and can play roles in genome stability, nuclear domain organization, and gene regulation. In some embodiments, the plurality of RNA transcripts comprises IncRNAs.

[0058] Small interfering RNAs (siRNAs) are short RNA molecules, typically 20-24 nucleotides long, which operate within the RNA interference (RNAi) pathway in plants. siRNA loci in the genome are transcribed to produce a double- stranded transcript, which is processed by enzymes within the cell such as Dicer or Dicer-like proteins to produce a mature siRNA. siRNAs may be involved in transcriptional gene silencing (TGS) or post- transcriptional gene silencing (PTGS) in plants. In some embodiments, the plurality of RNA transcripts comprises siRNAs.

[0059] MicroRNAs (miRNAs) are a class of small RNAs that bind to complementary regions of mRNAs, resulting in post-transcriptional gene silencing (PTGS). miRNAs are known for their role in development, abiotic stress response, and response to nutrient availability (See Millar (2020) Plants (Basel) 9(2): 198). In some embodiments, the plurality of RNA transcripts comprises miRNAs.

[0060] In some embodiments, the plurality of RNA transcripts comprises transcripts associated with yield, overall fitness, biomass, photosynthetic efficiency, nutrient use efficiency, heat tolerance, drought tolerance, herbicide tolerance, or disease resistance.

[0061] In some embodiments, the plurality of RNA transcripts comprises transcripts associated with flowering. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with flowering, floral transition, or flowering time, including but not limited to the following transcripts or orthologs thereof: CONSTANS (CO), SUPPRESSOR OF OVEREXPRESSION OF CONSTANS (SOC1), LEAFY (LF), TERMINAL FLOWER 1 (TFL1), FLOWERING LOCUS T (FT), FLOWERING LOCUS C (FLC), FRIGIDA (FRI), COOLAIR, COLD AIR, LONG HYPOCOTYL (HY), CASEIN KINASE 2 (CK2), microRNA156, and / or SQUAMOSA PROMOTER BINDING-LIKE (SPL) (See Freytes (2021) Curr Opin Plant Biol 63: 102049 for review).

[0062] In some embodiments, the plurality of RNA transcripts comprises transcripts associated with growth. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with cell division, including but not limited to transcripts of cyclinsand / or cyclin-dependent kinases. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with cell expansion. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with nutrient sensing or affected by the presence of nutrients and / or sugars, including but not limited to SNF1-RELATED PROTEIN KINASE (SNRK1), TARGET OF RAPAMYCIN (TOR), phytochrome interacting factors (PIFs), microRNA 156, and SQUAMOSA promoter binding protein-like (SPL) transcription factors. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with phytohormones. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with auxin, including but not limited to transcripts associated with auxin synthesis and transcripts associated with auxin response factors (ARFs); see Lastdrager et al. (2014) J Exp Bot. 65(3): 799-807. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with ethylene, including but not limited to transcripts associated with ethylene biosynthesis and transcripts associated with ethylene response factors (ERFs); see Dubois et al. (2018) Trends Plants Sci 23(4): 311-323. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with abscisic acid, including but not limited to transcripts associated with abscisic acid biosynthesis and transcripts associated with abscisic acid signaling; see Brookbank et al. (2021) Genes (Basel) 12(12): 1936.

[0063] In some embodiments, the plurality of RNA transcripts comprises transcripts associated with biotic stress resistance. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with biotic stress resistance and / or biotic stress response. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with resistance to herbivory, bacteria, fungi, viruses, nematodes, and / or insects. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with plant response to herbivory, bacteria, fungi, viruses, nematodes, and / or insects. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with phytohormones associated with response to biotic stress. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with jasmonic acid, including but not limited to transcripts associated with jasmonic acid biosynthesis and transcripts associated with jasmonic acid signaling. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with synthesis of secondary metabolites. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with the synthesis of defensive secondary metabolites.

[0064] In some embodiments, the plurality of RNA transcripts comprises transcripts associated with abiotic stress resistance. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with response to abiotic stressors including but not limited to drought, salinity, heat, cold, pH, and flooding. In some embodiments, the plurality of RNA transcripts comprises transcripts associated with phytohormones associated with abiotic stress responses, including but not limited to abscisic acid, auxin, brassinosteroids, cytokinins, ethylene, and gibberellins; see Waadt et al. (2022) Nat Rev Mol Cell Biol. 23(10): 680-694 for review; see Zhang et al. (2022) Nat Rev Genet 23(2): 104-119 for review.Outlier Score

[0065] In some embodiments, the expression level of a transcript in a plant or plant population of interest in comparison to the expression level of the transcript in one or more control plants or control plant populations may be represented in a numerical or mathematical form, such as an outlier score. The term “outlier score” refers to a representation of the degree of deviation of a transcript expression level in comparison to one or more reference transcript expression levels. The representation may be interpretable to humans and / or machines. In some embodiments, the reference transcript expression level is the mean expression level of a transcript within a population. In some embodiments, the reference transcript expression level is the median expression level of a transcript within a population. In some embodiments, the population of the reference transcript expression level is composed of transcript levels of the same locus in plants of the same species as the plant or plant population. In some embodiments, the population of the reference transcript expression level is composed of transcript levels of the same locus in plants of different cultivars of the same species as the plant or plant population. In some embodiments, the population of the reference transcript expression level of composed of transcript levels of orthologous loci in plants of one or more different species from the plant or plant population. In some embodiments, the population of the reference transcript expression level is composed of transcript levels of the same locus in one or more plants of the same species as the plant or plant population of interest. In some embodiments, the population of the reference transcript expression level is composed of transcript levels of the same locus in plants of different cultivars of the same species as the plant or plant population of interest. In some embodiments, the population of the reference transcript expression level of composed of transcript levels of orthologous loci in plants of one or more different species from the plantor plant population of interest. In some embodiments, the population of the reference transcript expression level is composed of transcript levels of paralogous genes.

[0066] In some embodiments, the outlier score is represented by the deviation from zero, wherein zero represents the reference transcript expression level. Outlier scores farther from zero indicate RNA transcript expression levels farther from the reference transcript expression level. In some embodiments, outlier scores do not indicate whether an RNA transcript expression level is increased or reduced in comparison to the reference transcript expression level. In other embodiments, a negative outlier score represents reduced expression of the RNA transcript expression level as compared to the reference transcript expression level, and a positive outlier score represents increased expression of the RNA transcript expression level as compared to the reference transcript expression level.

[0067] In some embodiments, the outlier score is expressed as the deviation of the mean scaled by the standard deviation of the population. The outlier score of an RNA transcript expression level can be assessed by quantifying standard deviation (s.d.) from the mean of the population or s.d. from the measured amount in an ancestral population or set of related species. In some embodiments, the outlier score is a Z-score. A Z-score expresses the deviation of a sample in a population from the population mean and is scaled by the standard deviation of the trait / gene in the population; thus, it is a metric that describes for an individual sample in a population the number of standard deviations that individual is from the mean. This approach is widely used in statistics and is most often applied to random variables that follow a Normal / Gaussian distribution; however, similar statistics can be derived for random variables that are not drawn from a normal distribution. In some embodiments, the outlier score is a non-parametric variant of a Z-score. A non-parametric variant of the Z-score can be applied by determining how far each sample deviates from the median of the distribution and scaling this value by the median of the absolute deviation of each sample from the median. In an exemplary embodiment, for an individual i, the Z-score can be calculated as Zi = (M(x) - Xi) / M(|M(x) - x|), where M is the median, M(x) indicates the median of the random vector x in the population, and Xi is the value for individual i in the population. In addition to non-parametric Z scores, Z-scores for other distributions can be derived by provided estimates of the dispersal, or by the defined point mass of the distribution.

[0068] In some embodiments, the outlier score is a burden score tied to a Z-score. A burden score can be computed by summing the number of genes that qualify as an outlier based on a Z-score approach for each individual in the population.

[0069] In some embodiments, the outlier score is determined by autoencoder-based outlier detection methods. As used herein, an “autoencoder” refers to a machine-learning technique that can be used for non-linear dimensionality reduction or compression of highdimensional data. In some embodiments, the autoencoder takes an input. In some embodiments, the autoencoder takes an input sample. An input sample is represented by an m-dimensional vector x. In some embodiments the input sample comprises expression measurements. The m-dimensional vector x is passed through a first series of one or more non-linear transformations, each of which can reduce the dimensionality such that the output vector of each transformation has fewer dimensions than the input vector. This first series of non-linear transformations is called the Encoder (E), and the output of the Encoder is an n- dimensional vector z = E(x), called a latent vector, where n < m. In some embodiments, the Encoder is an artificial neural network (ANN). In some embodiments, the Encoder comprises an artificial neural network (ANN). The latent vector z is then passed through a second series of non-linear transformations, each of which increases the dimensionality such that the output vector of each transformation has a higher dimension than the input vector. This second series of non-linear transformations, is called the Decoder (D), and the output of the decoder is an m-dimensional vector x’ = D(z), the same dimension as the input data point x. In some embodiments, the Decoder is an artificial neural network (ANN). In some embodiments, the Decoder comprises an artificial neural network (ANN). In some embodiments, both the Encoder and the Decoder are artificial neural networks. In some embodiments, both the Encoder and the Decoder comprise artificial neural networks. In some embodiments, a series of non-linear transformations comprises one or more non-linear transformations. The full autoencoder model is given by the composition of the Encoder and the Decoder, such that x’ = D(z) = D(E(x)). In some embodiments, the non-linear transformations of one of, or both of the Encoder transformations and the Decoder transformations can be modified by tuning a set of parameters 9. The autoencoder is trained by passing each sample through the model and iteratively tuning the parameters 9 via backpropagation through the series of non-linear transformations so that the output vector x’ is as similar to the input vector x as possible by minimizing a differentiable reconstruction error function e(x’, x). In some embodiments, the differentiable reconstruction error function is the mean squared error (MSE). Without wishing to be bound by theory, according to the low-rank hypothesis (also called the manifold hypothesis; See Cayton (2005) Univ, of California at San Diego Tech. Rep, 12(1— 17); Thibeault et al. (2022) arXiv preprint arXiv:2208.04848.) the high-dimensional data generated by complex physical systems tends to lie on a lower dimensional submanifold. Ineffect, the autoencoder learns a mapping between the high-dimensional measurement space X and the lower dimensional manifold Z. Under this assumption, samples that are representative of the underlying physical process will be compressed and reconstructed accurately by the autoencoder, while anomalous samples will incur a high reconstruction error. In some embodiments, the outlier score is determined by anomaly detection with attributed networks. This can be beneficial in situations where there is known structure among the samples in a dataset. In one non-limiting example, the samples in the dataset are measurements of RNA expression levels from a set of genes, and the known structure is a gene regulatory network comprising the set of genes. The autoencoder method can be modified to leverage the known structure as prior knowledge. To do this, each sample is represented as a node in a network. Nodes are connected by edges according to the network structure, and each node is assigned a vector of attributes x. In some embodiments, the vector of attributes x comprises measured expression values for a single gene in multiple tissues. In some embodiments, the Encoder (E) transformation is then constructed as a graph neural network (GNN), which aggregates attribute information across the network topology and outputs a lower-dimensional, compressed representation z of the attributes of each sample. Similarly, in some embodiments the Decoder (D) transformation is constructed as a graph neural network (GNN), which learns the inverse mapping from latent vectors z back to the original representation x. The graph autoencoder is given by the composition of the Encoder and the Decoder, such that x’ = D(z) = D(E(x)). In some embodiments, the non-linear transformations comprising one or both of the Encoder and Decoder GNNs can be modified by tuning a set of parameters 9. The graph autoencoder is trained by passing the full network of samples (as training data) through the model and tuning the parameters 9 via backpropagation through the GNN so that the output vector x’ is as similar to the input vector x as possible by minimizing a differentiable reconstruction error function e(x’, x). In some embodiments, the differentiable reconstruction error function is the mean squared error (MSE). Without wishing to be bound by theory, according to the low-rank hypothesis (also called the manifold hypothesis; See Cayton (2005) Univ, of California at San Diego Tech. Rep, 12(1-17); Thibeault et al. (2022) arXiv preprint arXiv:2208.04848) the high-dimensional data generated by complex physical systems tends to lie on a lower dimensional submanifold. In effect, the autoencoder learns a mapping between the high-dimensional measurement space X and the lower dimensional manifold Z. Under this assumption, samples that are representative of the underlying physical process will be compressed and reconstructed accurately by the autoencoder, while anomalous samples will incur a higher reconstruction error (e.g., above a threshold).

[0070] In some embodiments, the outlier score is determined by multivariate outlier detection methods wherein information is borrowed across genes in a local network and across tissues. In one embodiment, the Mahalanobis distance is used for multivariate outlier detection. The Mahalanobis distance measures the distance between a data point and the centroid of a multivariate distribution, taking into account the distribution of the data points for detecting outliers. In one non-limiting example, for gene expression data, this can be the deviation of the expression level for a gene in a particular sample relative to the mean of the population, while accounting for the covariance structure of the gene expression levels across the population. In some embodiments, Mahalanobis distance can be calculated using the following formula: Mahalanobis distance = sqrt((x-u) *A(-1) * (x-u)T), where x represents the gene expression vector of a sample, u represents the mean expression vector,A(-1) represents the inverse of the covariance or correlation matrix, and T represents the transpose operation used to transpose the vector (x-u) to ensure proper matrix multiplication between the vectors and matrices involved in the formula.

[0071] In some embodiments, the outlier score is determined using protein expression data in addition to RNA expression data. Additionally, characteristics or family-membership of genes can be used to classify those that are likely to show outlier status at the protein abundance level if the mRNA abundance level is classified as an outlier. Classification of those genes which are likely to show this correlation between mRNA and protein can potentially be determined by gene ontology terms or by the position of the gene within high dimensional embedding space (e.g., above a threshold) from a large protein neural network model such as a transformer.

[0072] In some embodiments, the outlier score is determined using splice isoform ratios. In some embodiments, all splice isoforms are considered together for a single RNA locus. In other embodiments, each splice isoform is considered individually for a single RNA locus.

[0073] In some embodiments, outlier scores for a single locus are determined at a tissuespecific level. In some embodiments, outlier status across multiple tissues can be aggregated to determine the level of confidence in calling an RNA transcript as beyond an outlier score threshold.

[0074] In some embodiments, a threshold is set to determine which loci are of further interest for selection or genetic modification. In some embodiments, the threshold is a one- tailed threshold on the distribution of outlier scores. In other embodiments, the threshold is a two-tailed threshold on the distribution of outlier scores. In some embodiments, the threshold is a set number of genes at one or more tails of the distribution of outlier scores. In someembodiments, the threshold is set to include the 10, 20, 30, 40, 50, 60, 70, 80 loci with the most extreme outlier scores. In some embodiments, the threshold is a discrete value for loci with the highest outlier scores. In some embodiments, the threshold is set to include the 10, 20, 30, 40, 50, 60, 70, 80 loci with the highest outlier scores. In some embodiments, the threshold is set to include the 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15% of loci with the most extreme outlier scores. In some embodiments, the threshold is set to include the 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15% of loci with the highest outlier scores. In some embodiments, the threshold is determined by dividing the number of transcripts that can be edited in one plant by the total number of transcripts in the plant. In other embodiments, the threshold is a p-value. In some embodiments, the threshold is set at three standard deviations from the mean. In some embodiments, the threshold is set at three standard deviations above the mean.Using outlier scores in breeding

[0075] Outlier scores can be leveraged to select plants to breed or hybridize. Crossing two plants generally results in an increase in the degree of genomic heterozygosity of the hybrid offspring, as loci that are homozygous for one allele in one parent may not be homozygous for that allele in the other parent. For example, a first inbred plant may have the genotype AA and a high outlier score at a locus, and a second inbred plant may have the genotype aa and a low outlier score at the same locus; crossing these plants results in a hybrid offspring with genotype Aa at this locus, most likely with an outlier score between that of the first inbred plant and the second inbred plant at the locus. Plants may be selected to breed or hybridize in order to reduce the outlier scores of one or more transcripts in the hybrid offspring. Plants may be selected to breed or hybridize in order to minimize the number of transcripts in the offspring with outlier scores exceeding a set threshold.

[0076] Also provided is a method for selecting two plant individuals to breed or hybridize, comprising: a. providing RNA expression data for a plurality of RNA transcripts of more than one plant or plant population; b. determining an outlier score for the expression level of each RNA transcript of the plurality of RNA transcripts in comparison to a one or more reference transcript expression levels for each of the more than one plants or plant populations; c. obtaining a distribution of the outlier scores for each of the more than one plants or plant populations; d. identifying in each of the more than one plants or plant populations a subset of RNA transcripts with outlier scores beyond a specified threshold in the distribution; and e. selecting two plant individuals capable of breeding or hybridizing withone another, wherein the two plant individuals have minimal overlap in their two subsets of RNA transcripts, or wherein the two plant individuals are derived from two plant populations with minimal overlap in their two subsets of RNA transcripts. Also provided is a method for selecting two plant individuals to breed or hybridize, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of more than one plant or plant population; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in each of the more than one plants or plant populations in comparison to a one or more reference transcript expression levels for each of the more than one plants or plant populations; c) obtaining a distribution of the plurality of outlier scores for each of the more than one plants or plant populations; d) identifying in each of the more than one plants or plant populations a subset of the plurality of RNA transcripts with outlier scores beyond a specified threshold in the distribution; and e) selecting two plant individuals capable of breeding or hybridizing with one another, wherein the two plant individuals have minimal overlap in their two subsets of RNA transcripts, or wherein the two plant individuals are derived from two plant populations with minimal overlap in their two subsets of RNA transcripts. In some embodiments, the overlap in the subsets of RNA transcripts from the two plant individuals is the set of RNA transcripts beyond the set threshold in both of the two plants. In some embodiments, minimal overlap is defined as one or more RNA transcripts present in the subset of RNA transcripts of the first of the two selected plant individuals are not present in the subset of RNA transcripts of the second of the two selected plant individuals. In some embodiments, minimal overlap is defined as between two and twenty RNA transcripts present in the subset of RNA transcripts of the first of the two selected plant individuals are not present in the subset of RNA transcripts of the second of the two selected plant individuals. In some embodiments, minimal overlap is defined as fewer than 30, 25, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, or 2 of the RNA transcripts beyond the set threshold of the first of the two selected plant individuals are also present in the subset of RNA transcripts beyond the set threshold of the second of the two selected plants. In some embodiments, minimal overlap is defined as less than 50%, 40%, 30%, 25%, 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, of the RNA transcripts beyond the set threshold of the first of the two selected plant individuals are also present in the subset of RNA transcripts beyond the set threshold of the second of the two selected plants. 1

[0077] Machine learning algorithms can be utilized to improve methods of plant breeding. For example, machine learning algorithms may automatically and quickly produce desired results like identifying, quantifying, and / or characterizing putative pairs of plants for breeding, using information from outlier scores for loci in all putative parent plants. In some embodiments, machine learning algorithms may be more robust than conventional algorithms. For example, a conventional algorithm may only be capable of identifying, quantifying, and / or characterizing pairs of plants in ways that a programmer has specifically accounted for. A machine learning algorithm may be able to process data with unanticipated characteristics (e.g., sequence-based, expression-based, or epigenetics -based signatures) based on training data provided to the machine learning algorithm. In some embodiments, machine learning algorithms may be more accurate than conventional algorithms. For example, a machine learning algorithm may improve over time with reinforcement mechanisms (e.g., if a human operator periodically performs manual checks on machine learning outputs). In some embodiments, conventional algorithms may have to be re-written and / or modified to improve the algorithm’s accuracy.

[0078] In some embodiments, algorithms are used which seek to minimize homozygous haplotypes that lead to aberrant expression, considering thousands of loci genome-wide based on what would arise or be segregated out to the next generation during meiosis. Some such algorithms are known in the art (See Jannink et al. (2010) Briefings in Functional Genomics 9(2): 166-177; Heffner et al. (2010) Crop Science 50(5): 1681-1690; Meuwissen et al. (2001) Genetics 157(4): 1819-1829).

[0079] In some embodiments, any of the machine-learning models described herein can be trained and re-trained iteratively over time. For example, the autoencoder model may be trained using a first set of training data during a first time period, and then retrained using a second set of training data during a second time period. The second set of training data may include new training data (e.g., obtained during the deployment of the model after it is trained using the first set of training data) such that the model can be iteratively trained via a feedback loop.Genetic Modifications

[0080] In some embodiments of the methods provided herein, one or more genetic modifications are determined for each RNA transcript of the subset of RNA transcripts with outlier scores beyond a threshold in the distribution of outlier scores, wherein the determinedgenetic modifications are designed to change the RNA expression of the RNA transcript to have a reduced magnitude of outlier score.

[0081] The genetic modifications of the methods herein may be made to one or more genomic regions selected independently from the group consisting of a gene, a promoter, an enhancer, a transcription factor binding site, a splice donor site, a splice acceptor site, an miRNA binding site, an siRNA binding site, an insulator, an intron, an exon, a 5’ UTR, a 3’ UTR, a suppressor, a termination sequence, a sequence encoding a non-coding RNA, a transgene, and a quantitative trait locus (QTL). In some embodiments, the genetic modification results in the insertion or deletion of nucleotides at or near the RNA transcript locus sequence. In some embodiments, the genetic modification results in an insertion of at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, or 40 nucleotides at or near the RNA transcript locus sequence. In other embodiments, the genetic modification determined for an individual RNA transcript is a small element insertion to increase expression. As used herein, a “small element insertion” refers to an insertion of no more than 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, or 40 nucleotides. In some embodiments, the genetic modification results in a deletion of at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 60, 70, 80, 90, 100, 150, 200, 250, 300, 350, 400, 450, 500, 550, 600, 650, 700, 750, 800, 850, 900, 950, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 12500, 15000, 17500, 20000, 22500, or 25000 nucleotides at or near the RNA transcript locus sequence. In some embodiments, the genetic modification determined for an individual RNA transcript is a small deletion in the promoter. As used herein, a “small deletion” refers to a deletion of no more than 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 60, 70, 80, 90, or 100 nucleotides. In some embodiments, the genetic modification results in a nucleotide substitution at or near the RNA transcript locus sequence. In some embodiments, the genetic modification results in a substitution of at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 60, 70, 80, 90, or 100 nucleotides at or near the RNA transcript locus sequence.

[0082] In some embodiments, the genetic modification results in the formation of a novel sequence at the locus. In some embodiments, the genetic modification results in the formation of a novel allele of a gene. In other embodiments, the genetic modification results in the formation of an alternative allele of a gene that was not present in the plant before the modification. In some embodiments, the alternative allele is an allele present in the plant from which the reference transcript expression level is derived. In some embodiments, the alternative allele is an allele from a different plant population. In some embodiments, thealternative allele is an allele from a different cultivar. In some embodiments, the alternative allele is an allele from a different species of plant. In some embodiments, the alternative allele is an allele with a reduced outlier score from a different plant.

[0083] In some embodiments, the genetic modification reduces the outlier score, i.e., the difference between the RNA expression level of the transcript in the modified plant cells and the reference transcript expression level is less than the difference between the RNA expression level of the transcript in the unmodified plant cells and the reference transcript expression level. In some embodiments, the genetic modification reduces the outlier score such that the outlier score is no longer beyond the designated threshold. In some embodiments, the genetic modification reduces the outlier score by at least one standard deviation of the population of RNA transcript expression levels.

[0084] In some embodiments, multiple modifications are determined for a single RNA transcript locus. In other embodiments, a single modification is determined for a single RNA transcript locus. In some embodiments, one or more modifications may be determined for or within each RNA transcript locus.

[0085] The method of the present invention may be used to introduce edits to affect any phenotype, quality, or trait of the organism. For instance, the methods herein may be used to introduce edits to the genome that affect yield, overall fitness, biomass, photosynthetic efficiency, nutrient use efficiency, heat tolerance, drought tolerance, herbicide tolerance, or disease resistance of a plant.

[0086] Artificial intelligence and / or machine learning may be used to perform the methods herein. Machine learning algorithms can be advantageous for a number of reasons. For example, machine learning algorithms may automatically and quickly produce desired results like identifying, quantifying, and / or characterizing putative target sites for editing. In some embodiments, machine learning algorithms may be more robust than conventional algorithms. For example, a conventional algorithm may only be capable of identifying, quantifying, and / or characterizing target sites in ways that a programmer has specifically accounted for. A machine learning algorithm may be able to process data with unanticipated characteristics (e.g., sequence-based, expression-based, or epigenetics -based signatures) based on training data provided to the machine learning algorithm. In some embodiments, machine learning algorithms may be more accurate than conventional algorithms. For example, a machine learning algorithm may improve over time with reinforcement mechanisms (e.g., if a human operator periodically performs manual checks on machinelearning outputs). In some embodiments, conventional algorithms may have to be re-written and / or modified to improve the algorithm’s accuracy.

[0087] The genetic modifications of the present invention may be determined, assessed, weighted, or prioritized by a statistical model based on one or more criteria. Examples of the criteria include, but are not limited to, evolutionary conservation (See e.g. Chun and Fay (2009) Genome R es. 19: 1553-1561 and Rodgers-Melnick et al (2015) PNAS 112: 3823- 3828), functional impact of amino acid change (See e.g. Ng et al (2003) NAR 31: 3812-3814 and Adzhubei et al (2010) Nat Methods 7:248-249), functional impact of protein conformation and / or stability (See e.g. Rosetta, a computational protein design platform from Cyrus Bio Inc.), adjacency to selective sweep regions (See e.g. Hufford et al (2012) Nat gen 44: 808-813), and outlier status of an endophenotype (See e.g. Zhao et al (2016) AJHG 98, 299-309). In some embodiments, the evolutionary conservation is determined by sequence alignment in a genic or an intergenic region. In some embodiments, the functional impact of amino acid change is weighted according to the blocks substitution matrix (BLOSUM). In some embodiments, the functional impact of protein conformation and / or stability is determined by a Monte Carlo search for minimal free energy. In some embodiments, the functional impact of protein conformation and / or stability is predicted by learning a representation of amino acid order from existing proteins in higher dimensional space. In some embodiments, the selective sweep region is determined by a decrease of pairwise nucleotide diversity 7t or linkage disequilibrium relative to the rest of the genome. In some embodiments that may be combined with any of the preceding embodiments, the feature is a numeric or categorical value associated with a specific allele at a genomic locus.Genome Editing

[0088] In some embodiments that may be combined with any of the preceding embodiments, genomic alteration is achieved by genome editing. As used herein, the term “genome editing” or “gene editing” refers to the process of altering the target genomic DNA sequence by inserting, replacing, or removing one or more nucleotides.

[0089] Genome editing may be accomplished by using nucleases, which create specific double-strand breaks (DSBs) at desired locations in the genome and harness the endogenous mechanisms of the cell to repair the induced break by homology-directed repair (HDR) (e.g., homologous recombination) or by non-homologous end joining (NHEJ). Any suitable nuclease may be introduced into a cell to induce genome editing of a target DNA sequence including, but not limited to, clustered regularly interspersed short palindromic repeats(CRISPR)-associated protein (Cas, e.g. Cas9 and Casl2a) nucleases, zinc finger nucleases (ZFNs, e.g. FokI), transcription activator-like effector nucleases (TALENs, e.g. TALEs), meganucleases, and variants thereof (Shukla et al. (2009) Nature 459: 437- 441 ; Townsend et al (2009) Nature 459: 442-445). Accordingly, in some embodiments of the present invention, the genome editing is achieved by a clustered regularly interspersed short palindromic repeats (CRISPR) system, a transcription activator-like effector nuclease (TALEN) system, or a zinc finger nuclease (ZFN) system.

[0090] CRISPR technology for editing the genes of eukaryotes is disclosed in the art, such as in U.S. Patent Application Publication US2015 / 0344912A1, and in U.S. Pat. Nos. 8,697,359, 8,771,945, 8,945,839, 8,999,641, 8,993,233, 8,895,308, 8,865,406, 8,889,418, 8,871,445, 8,889,356, 8,932,814, 8,795,965, 8,906,616, and 10,227,611. Cpfl (Casl2a) endonucleases and corresponding guide RNAs and PAM sites are disclosed in U.S. Pat. No. 9,790,490 and U.S. patent application Ser. No. 15 / 566,528 (national phase of PCT Application PCT / EP2016 / 058442, published as WO 2016 / 166340), now published as U.S. Patent Application Publication 2018 / 0282713. Other CRISPR nucleases useful for editing genomes include C2cl and C2c3 (see Shmakov et al. Mol. Cell 2015, 60: 385-397), CasX and CasY (see Burstein et al. Nature 2016, doi:10.1038 / nature21059), and Casl2j (see WO 2021 / 216512). Plant RNA promoters for expressing CRISPR guide RNA and plant codon- optimized CRISPR Cas9 endonuclease are disclosed in U.S. patent application Ser. No. 15 / 120,110, published as U.S. Patent Application Publication 2017 / 0166912, national phase application claiming priority to International Patent Application PCT / US2015 / 018104 (published as WO 2015 / 131101 and claiming priority to U.S. Provisional Patent Application 61 / 945,700). Methods of using CRISPR technology for genome editing in plants are disclosed in in U.S. Patent Application Publications U.S. 2015 / 0082478 Al and U.S.2015 / 0059010A1 and in International Patent Application PCT / US2015 / 038767 Al (published as WO 2016 / 007347, claiming priority to U.S. Provisional Patent Application 62 / 023,246, with U.S. National Phase application U.S. Ser. No. 15 / 325,116, now published as U.S. Patent Application Publication 2017 / 0306349).

[0091] In some embodiments, the type of genome editing is base editing. As used herein, the term "base editing" refers to a base mutation (substitution, deletion, or addition) that causes point mutations in a target site within a target gene, with a few bases (one or two). Various base editors are known in the art and may have various approximate editing windows. See e.g., Rees, H.A. and Liu, D.R., 2018. Base editing: precision chemistry on the genome and transcriptome of living cells. Nature reviews genetics, 79(12), pp.770-788;Molla, K.A. and Yang, Y., 2019. CRISPR / Cas-mediated base editing: technical considerations and practical applications. Trends in biotechnology, 37(10), pp.1121-1142; and Mishra, R., Joshi, R.K. and Zhao, K., 2020. Base editing in crops: current advances, limitations, and future implications. Plant Biotechnology Journal, 18 V), pp.20-31. Accordingly, in some embodiments, the editing window is from 5-10 bp. In some embodiments, the editing window is from 5-15 bp. In some embodiments, the editing window is from 5-20 bp. In some embodiments, the editing window is from 5-25 bp. In some embodiments, the editing window is from 5-30 bp. In some embodiments, the editing window is from 5-35 bp. In some embodiments, the editing window is from 5-40 bp. In some embodiments, the editing window is from 5-45 bp. In some embodiments, the editing window is from 5-50 bp. In some embodiments, the editing window is from 10-20 bp. In some embodiments, the editing window is from 10-30 bp. In some embodiments, the editing window is from 10-40 bp. In some embodiments, the editing window is from 10-50 bp.

[0092] Desired DNA sequence modifications can be accomplished through the use of prime editing (Anzalone et al. Nature 2019, 576(7785): 149-157). In some embodiments, prime editing uses (i) a Cas nickase, in some embodiments a Cas9 nickase, in other embodiments a Cas 12 nickase, fused to a reverse transcriptase (nCas-RT), in some embodiments an M-MLV reverse transcriptase, and (ii) a prime editing Cas guide RNA (pegRNA) that both specifies the genome target site and has an extension that encodes the target edit within a template for the reverse transcriptase. The binding of the pegRNA directs the Cas nickase to create a single-stranded break in the DNA at the nicking site. The extension of the pegRNA binds to the nicked DNA that has an exposed 3 ’-hydroxyl group, priming the reverse transcriptase to produce a DNA strand that is complementary to the extension of the pegRNA. This DNA strand will include the complement to any desired edits present in the provided pegRNA extension. Mismatch repair by the cell will then resolve the mismatch between the unedited parent strand and the edited product of the reverse transcriptase, thus introducing the desired edits into the genome. Prime editing systems may also include elements to inhibit mismatch repair, or to nick the unedited parent strand to increase editing efficiency. A mobility element can be fused to the pegRNA so as not to interfere with priming of the reverse transcriptase.

[0093] In some embodiments, prime editing can also be accomplished with Cas nucleases in place of Cas nickases (see Adikusuma et al. Nucleic Acids Res. 2021, 49(18): 10785-10795). In some embodiments, prime editing uses (i) a Cas nuclease, in some embodiments a Cas9 nuclease, in other embodiments a Cas 12 nuclease, fused to a reverse transcriptase (Cas-RT), in some embodiments a M-MLV reverse transcriptase, and (ii) a prime editing Cas guide RNA (pegRNA) that both specifies the genome target site and has an extension that encodes the target edit within a template for the reverse transcriptase. In some embodiments, the binding of the pegRNA directs the Cas nuclease to create a double- stranded break in the DNA at the target site. The extension of the pegRNA binds to the cut DNA that has an exposed 3 ’-hydroxyl group, priming the reverse transcriptase to produce a DNA strand that is complementary to the extension of the pegRNA. This DNA strand will include the complement to any desired edits present in the provided pegRNA extension. Mismatch repair by the cell will then resolve the mismatch between the unedited parent strand and the edited product of the reverse transcriptase, thus introducing the desired edits into the genome. Prime editing systems may also include elements to inhibit mismatch repair, or to nick the unedited parent strand to increase editing efficiency. A mobility element can be fused to the pegRNA so as not to interfere with priming of the reverse transcriptase.

[0094] Prime editing makes precise DNA sequence modifications rather than random insertions, deletions, and substitutions (Indels), thus increasing the probability of obtaining the desired effect. Prime editing may be used to introduce any single base pair substitution as well as small deletion or insertions. Deletions of up to 80 base pairs have been produced using prime editing with a single pegRNA in human cells, and insertions of up to 40 base pairs (see Anzalone et al. Nature 2019, 576: 149-157). Dual pegRNA systems are also known in the art (see Choi et al. Nat Biotechnol 2021, 40(2): 218-226; Lin et al. Nature Biotechnology 2021, 39(8): 923-927) and can be used to generate precise large deletions, or to improve editing efficiency for small insertions, deletions, or substitutions. Additionally, dual pegRNA systems where the extension of the pegRNAs is not complementary to the endogenous locus, but are complementary to one another, can be used to replace endogenous sequence and / or mediate larger insertions (see Anzalone et al. Nat Biotechnol 2022, 40(5): 731-740).

[0095] The determined genetic modifications alone or in combination may be tested or validated to analyze the effect of the one or more genetic modifications on the outlier status of one or more transcripts. In some embodiments, validation analysis is performed in vitro. In some embodiments, validation analysis is performed in silico. In some embodiments, validation analysis is performed in cell culture. In some embodiments, validation analysis is performed in protoplasts. In some embodiments, validation of edits is performed in a high- throughput manner. In some embodiments, validation is performed by Perturb-Seq (aka CRISP-seq, aka CROP-Seq), a method of single-cell RNA sequencing after pooled geneticperturbation screens (see Dixit et al. (2016) Cell 167(7): 1853-1866, el7). In some embodiments, validation analysis consists of making one or more genetic modifications to one or more plant cells, then extracting RNA from both modified and unmodified plant cells by methods known in the art, and finally comparing the RNA expression level of the transcript in the modified plant cells with the RNA expression level of the transcript in the unmodified plants cells by methods known in the art. In some embodiments, alternative genetic modifications are determined before the validation analysis is performed. In other embodiments, alternative genetic modifications are determined after validation analysis is performed for a first determined genetic modification. In some embodiments, further validation analysis is performed on the alternative genetic modifications. In some embodiments, the determination of alternative genetic modifications and validation analysis of the genetic modifications is repeated until a genetic modification with a desired outcome is obtained. In some embodiments, the desired outcome is a reduction of the outlier score, i.e., the difference between the RNA expression level of the transcript in the modified plant cells and the reference transcript expression level is less than the difference between the RNA expression level of the transcript in the unmodified plant cells and the reference transcript expression level. In some embodiments, the desired outcome is a reduction of the outlier score such that the outlier score is no longer beyond the designated threshold.Plants

[0096] The plant of the present invention may be any plant that is of economic and / or scientific value to humans. Examples of plant species of interest include, but are not limited to, com (Zea mays'), Brassica sp. (e.g., B. napus, B. rapa, B. juncea), particularly those Brassica species useful as sources of seed oil, alfalfa (Medicago sativa), rice (Oryza sativa), rye (Secale cereale), sorghum (Sorghum bicolor, Sorghum vulgare), camelina (Camelina sativa), millet (e.g., pearl millet (Pennisetum glaucum), proso millet (Panicum miliaceum), foxtail millet (Setaria italica), finger millet (Eleusine coracana)), sunflower (Helianthus annuus), quinoa (Chenopodium quinoa), chicory (Cichorium intybus), lettuce (Lactuca sativa), safflower (Carthamus tinctorius), wheat (Triticum aestivum), soybean (Glycine max), tobacco (Nicotiana tabacum), potato (Solanum tuberosum), peanuts (Arachis hypogaea), cotton (Gossypium barbadense, Gossypium hirsutum), sweet potato (Ipomoea batatus), cassava (Manihot esculenta), coffee (Coffea spp.), coconut (Cocos nucifera), pineapple (Ananas comosus), citrus trees (Citrus spp.), cocoa (Theobroma cacao), tea (Camellia sinensis), banana (Musa spp.), avocado (Persea americana), fig (Ficus casica), guava(Psidium guajava), mango (Mangifera indica), olive Olea europaea), papaya (Carica papaya), cashew (Anacardium occidentale). macadamia (Macadamia inlegrifolia), almond (Prunus amygdalus), sugar beet (Beta vulgaris), sugarcane (Saccharum spp.), oil palm (Elaeis guineensis), poplar (Populus spp.), eucalyptus (Eucalyptus spp.), oats (Avena sativa), barley (Hordeum vulgare), sesame (Sesamum spp.), flax (Linum usitatissimum), cannabis (Cannabis spp.), a vegetable crop, a forage crop, an industrial crop, a woody crop, a biomass crop, an ornamental, and a conifer. In some embodiments, the plant is selected from the group consisting of corn, soy, and wheat.Performance

[0097] The performance of a plant may refer to any phenotype, quality, or trait of the plant. For instance, in some embodiments, the performance may be yield, overall fitness, biomass, photosynthetic efficiency, nutrient use efficiency, heat tolerance, drought tolerance, herbicide tolerance, or disease resistance. In some embodiments, the performance is yield performance in maize. “Yield performance” refers to the total amount of harvestable material, e.g., grain or forage, obtained in a typical field performance trial. In some embodiments that may be combined with any of the preceding embodiments, the performance is a quantitative trait.

[0098] In some embodiments, the methods of the present invention include comparing the performance of a plant comprising the determined genetic modifications to the performance of a control plant. In some embodiments, the control plant is an unmodified plant from the same population as the modified plant. In other embodiments, the control plant is a wildtype plant. In other embodiments, the control plant is an unmodified plant from another cultivar of the same species as the modified plant. In some embodiments, the control plant is an unmodified plant from the same species as the modified plant. In certain aspects, the present invention provides a plant with improved performance produced or selected by any one of the preceding methods.EMBODIMENTS1. A method for improving performance of a plant, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of the plant or of a plant population comprising the plant;b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in the plant or plant population in comparison to one or more reference transcript expression levels; c. obtaining a distribution of the plurality of outlier scores; d) identifying a subset of the plurality of RNA transcripts with the plurality of outlier scores beyond a threshold in the distribution; e) determining, for each RNA transcript of the subset of RNA transcripts, a genetic modification that changes the RNA expression of the RNA transcript to have a reduced outlier score magnitude; and f) introducing into the plant the genetic modifications determined for one or more of the RNA transcripts of the subset of RNA transcripts, thereby improving the performance of the plant.2. The method of embodiment 1, wherein the method further comprises the step of: g) comparing the performance of the plant comprising the determined genetic modifications to the performance of a control plant.3. A method of identifying targets for genetic modification to improve performance of a plant, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of the plant or of a plant population comprising the plant; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in the plant or plant population in comparison to one or more reference transcript expression levels; c) obtaining a distribution of the plurality of outlier scores; d) identifying a subset of the plurality of RNA transcripts with the plurality of outlier scores beyond a threshold in the distribution; and e) determining, for each RNA transcript of the subset of RNA transcripts, a genetic modification that changes the RNA expression of the RNA transcript to have a reduced outlier score magnitude, thereby identifying targets for genetic modification to improve plant performance.4. The method of embodiment 3, wherein the method further comprises the step of: f) selecting a subset of targets for genetic modification in a plant from the determined genetic modifications, wherein the selected genetic modifications arepredicted to reduce the magnitude of RNA expression outlier scores of the RNA transcripts.5. A method for selecting two plant individuals to breed or hybridize, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of more than one plant or plant population; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in each of the more than one plants or plant populations in comparison to one or more reference transcript expression levels; c) obtaining a distribution of the plurality of outlier scores for each of the more than one plants or plant populations; d) identifying in each of the more than one plants or plant populations a subset of the plurality of RNA transcripts with the plurality of outlier scores beyond a specified threshold in the distribution; and e) selecting two plant individuals capable of breeding or hybridizing with one another, wherein the two plant individuals have minimal overlap in their two subsets of RNA transcripts, or wherein the two plant individuals are derived from two plant populations with minimal overlap in their two subsets of RNA transcripts.6. The method of any one of embodiments 1-5, wherein the method further comprises steps e’), e”), and e’”) after step e) and before any following steps, comprising: e’) analyzing the effect of one or more genetic modifications determined in step e) on RNA expression; e”) determining one or more alternative genetic modifications for an RNA transcript and performing additional analysis; and e”’) selecting a genetic modification for each RNA transcript in the subset of RNA transcripts.7. The method of embodiment 6, wherein the analysis is performed in vitro, in silica, in cell culture, in protoplasts, and / or by Perturb-Seq.8. The method of embodiment 6, wherein step e”) is repeated until a genetic modification with a desired outcome is obtained.9. The method of embodiment 8, wherein the desired outcome is a reduction of the outlier score for the RNA transcript such that the outlier score is no longer beyond the threshold.10. The method of any one of embodiments 1-9, wherein the RNA expression levels are not pooled from multiple individuals.11. The method of any one of embodiments 1-9, wherein the RNA expression levels are combined from multiple individuals from the plant population.12. The method of any one of embodiments 1-11, wherein the outlier score is determined using RNA expression levels across multiple tissues.13. The method of embodiment 12, wherein the RNA expression levels are from a sample comprising multiple tissues, or are pooled from multiple tissues.14. The method of any one of embodiments 1-13, wherein the outlier scores are determined using RNA expression levels from a single tissue.15. The method of any one of embodiments 1-14, wherein the outlier score is determined using RNA expression levels across multiple members of a gene family.16. The method of any one of embodiments 1-14, wherein the outlier score is determined using RNA expression levels across multiple members of a gene interaction network.17. The method of any one of embodiments 1-16, wherein the outlier score is determined using protein expression levels in addition to RNA expression levels.18. The method of any one of embodiments 1-17, wherein the outlier score is determined using splice isoform ratios.19. The method of any one of embodiments 1-18, wherein the RNA expression levels are from whole genome RNA-seq data.20. The method of any one of embodiments 1-19, wherein the RNA expression levels are derived from a micro level tissue.21. The method of embodiment 20, wherein the micro level tissue is selected from the group consisting of shoot apical meristem, inflorescence meristem, immature male embryos, immature female embryos, developing anthers, and root hairs.22. The method of any one of embodiments 20 or 21, wherein the RNA expression levels are obtained by single cell sequencing.23. The method of any one of embodiments 20 or 21, wherein the micro level tissue is obtained by laser capture microdissection.24. The method of any one of embodiments 1-23, wherein the RNA expression levels are derived from tissue isolated at a specific plastochron.25. The method of any one of embodiments 1-19, wherein the RNA expression levels are derived from macro level tissue.26. The method of embodiment 25, wherein the macro level tissue is selected from the group consisting of leaf, root, internode, anther, tassel, ear shank, kernel, and ligule.27. The method of any one of embodiments 1-19, wherein the RNA expression levels are derived from protoplasts, from undifferentiated cells, from plant calli, or from plant cell suspension cultures.28. The method of any one of embodiments 1-27, wherein the plurality of RNA transcripts comprises mRNAs.29. The method of any one of embodiments 1-28, wherein the plurality of RNA transcripts comprises IncRNAs.30. The method of any one of embodiments 1-29, wherein the plurality of RNA transcripts comprises miRNAs.31. The method of any one of embodiments 1-30, wherein the plurality of RNA transcripts comprises siRNAs.32. The method of any one of embodiments 1-31, wherein the plurality of RNA transcripts comprises transcripts associated with flowering, transcripts associated with growth, transcripts associated with biotic stress resistance, and / or transcripts associated with abiotic stress resistance.33. The method of any one of embodiments 1-32, wherein the plurality of RNA transcripts consists of all RNA transcripts.34. The method of any one of embodiments 1-33, wherein the reference transcript expression level is the RNA expression level of the RNA transcript in another plant of the same species.35. The method of any one of embodiments 1-33, wherein the reference transcript expression level is the consensus transcript level within a population of plants of the same species.36. The method of any one of embodiments 1-33, wherein the reference transcript expression level is the consensus transcript level across multiple cultivars of the same species.37. The method of any one of embodiments 1-33, wherein the reference transcript expression level is the consensus transcript level of orthologs across multiple species.38. The method of any one of embodiments 1-37, wherein the outlier score is expressed as the deviation of the mean scaled by the standard deviation of the population.39. The method of any one of embodiments 1-37, wherein the outlier score is a Z-score.40. The method of any one of embodiments 1-37, wherein the outlier score is a burden score tied to a Z-score.41. The method of any one of embodiments 39 or 40, wherein the Z-score is a nonparametric Z-score.42. The method of any one of embodiments 39 or 40, wherein the Z-score is derived from provided estimates of the dispersal or from the defined point mass of the distribution.43. The method of any one of embodiments 1-37, wherein the outlier score is determined by autoencoder-based outlier detection, wherein the autoencoder-based outlier detection comprises an Encoder and a Decoder, wherein the Encoder and the Decoder each comprise one or more non-linear transformations.44. The method of embodiment 43, wherein the autoencoder takes an input, wherein the input to the autoencoder comprises RNA expression measurements.45. The method of any one of embodiments 43 or 44, wherein the Encoder of the autoencoder comprises an artificial neural network (ANN).46. The method of any one of embodiments 43-45, wherein the Decoder of the autoencoder comprises an artificial neural network (ANN).47. The method of any one of embodiments 43-46, wherein the outlier score is determined by anomaly detection with attributed networks.48. The method of embodiment 47, wherein the outlier score is determined for a set of genes in a known gene regulatory network.49. The method of any one of embodiments 47 or 48, wherein the autoencoder takes an input, wherein the input to the autoencoder comprises a vector comprising measured expression values for a single gene across multiple tissues.50. The method of any one of embodiments 47-49, wherein the Encoder is constructed as a graph neural network (GNN).51. The method of any one of embodiments 47-50, wherein the Decoder is constructed as a graph neural network (GNN).52. The method of any one of embodiments 43-51, wherein the non-linear transformations of the Encoder, the Decoder, or both the Encoder and Decoder are modified by tuning a set of parameters (9).53. The method of embodiment 52, wherein the parameters (9) are trained by passing an input vector through the model to produce an output vector, and iteratively tuning the parameters (9) via backpropagation through the series of non-linear transformations byminimizing a differentiable reconstruction error function between the output vector and the input vector.54. The method of embodiment 53, wherein the differentiable reconstruction error function is the mean squared error (MSE).55. The method of any one of embodiments 1-37, wherein the outlier score is determined by multivariate outlier detection methods, wherein information is borrowed across genes in a local network and across tissues.56. The method of embodiment 55, wherein the multivariate outlier detection methods comprise use of the Mahalanobis distance.57. The method of any one of embodiments 1-56, wherein the threshold is a one-tailed threshold on the distribution, a two-tailed threshold on the distribution, a p-value, or a set number of genes at one or more tails of the distribution.58. The method of embodiment 57, wherein the threshold is set at 50 genes at the upper tail of the distribution.59. The method of any one of embodiments 1-58, wherein the genetic modification determined for an individual RNA transcript creates a novel allele.60. The method of any one of embodiments 1-58, wherein the genetic modification determined for an individual RNA transcript replicates an alternative allele not present in the plant.61. The method of embodiment 60, wherein the alternative allele is an allele present in the plant from which the reference transcript expression level is derived.62. The method of any one of embodiments 60 or 61, wherein the alternative allele is an allele with a reduced outlier score from another plant.63. The method of any one of embodiments 1-62, wherein the genetic modification determined for an individual RNA transcript targets a DNA sequence element selected from the group consisting of a promoter, an enhancer, a transcription factor binding site, a splice donor site, a splice acceptor site, an miRNA binding site, an siRNA binding site, an insulator, an intron, an exon, a 5’ UTR, and a 3’ UTR.64. The method of embodiment 63, wherein the genetic modification determined for an individual RNA transcript is a small deletion in the promoter.65. The method of embodiment 63, wherein the genetic modification determined for an individual RNA transcript is a small element insertion to increase expression.66. The method of any one of embodiments 1-65, wherein the genome of the plant is modified using a nuclease selected from the group consisting of a Cas nuclease, a TALE nuclease, a zinc finger nuclease, a meganuclease, or an engineered nuclease.67. The method of embodiment 66, wherein the Cas nuclease is a Cas9 or Cas 12 nuclease.68. The method of any one of embodiments 1-67, wherein the reduced outlier score magnitude is no longer beyond the threshold, and / or is reduced by at least one standard deviation of the distribution.69. The method of any one of embodiments 1-68, wherein the plant is selected from the group consisting of corn, soy, and wheat.70. The method of any one of embodiments 1-69, wherein the RNA expression levels are obtained from a plant from an inbred panel, a hybrid panel, or an asexually-propagated population.71. The method of any one of embodiments 1-70, wherein improved performance of the plant is measured by a metric selected from the group consisting of growth, yield, survival, biomass, number of kernels, ear weight, and nutritional content.72. The method of any one of embodiments 5-71, wherein:(a) one or more RNA transcripts present in the subset of RNA transcripts of the first of the two selected plant individuals are not present in the subset of RNA transcripts of the second of the two selected plant individuals;(b) between two and twenty RNA transcripts present in the subset of RNA transcripts of the first of the two selected plant individuals are not present in the subset of RNA transcripts of the second of the two selected plant individuals;(c) the selected two plants capable of breeding or hybridizing with one another have less than 50% of the subset of RNA transcripts overlapping between their two subsets of RNA transcripts; and / or(d) the selected two plants capable of breeding or hybridizing with one another have no overlap in their two subsets of RNA transcripts.73. A plant produced by the method of any one of embodiments 1, 2, and 5-72.74. A seed produced by the method of any one of embodiments 5-72, or by the plant of embodiment 73.75. A plant grown from the seed of embodiment 74.EXAMPLES

[0099] The presently disclosed subject matter will be better understood by reference to the following Examples, which are provided as exemplary of the invention, and not by way of limitation.Example 1: Calculation of outlier scores based on RNA expression

[0100] RNAseq data including RNA expression levels are obtained from a representative individual from a first cultivar of plant. RNAseq data including RNA expression levels are also obtained from representative individuals from several other cultivars of the same species of plant. Gene expression is calculated for every gene in the genome for each sample. For every gene, a mean expression level and standard deviation are determined for the population of all cultivars sampled. An outlier score is calculated for each gene in the genome of the first cultivar, for example by determining the Z-score of that gene’s expression level, namely, how many standard deviations above the population mean the gene expression level is in the first cultivar. Alternatively, outlier scores are calculated using burden scores tied to Z-scores, autoencoder-based outlier detection methods, anomaly detection with attributed networks, or multivariate outlier detection methods wherein information is borrowed across genes in a local network and across tissues.Example 2: Determination of gene editing targets based on individual outlier scores

[0101] Outlier scores are calculated as in Example 1. The absolute values of the outlier scores for every gene in the first cultivar are ranked in order from greatest to least to form a distribution. A threshold is determined based on the number of genes to be edited and the number of genes in the genome. For instance, if it is determined that 50 genes out of 40,000 genes in the genome will be edited, the 50 genes with the highest outlier scores are selected for editing.Example 3: Determination of gene editing targets based on population outlier scores

[0102] Outlier scores are calculated as in Example 1. A threshold is determined based on the number of genes to be edited and the number of genes in the genome. For instance, if it is determined that roughly 50 genes out of 40,000 genes in the genome will be edited, a one- tailed threshold is set at 50 / 40,000, or 0.00125, or a two-tailed threshold is set at 0.00125 / 2,or 0.000625. All genes from the first cultivar with outlier scores correlated with exceeding this threshold in the distribution calculated in Example 1 are selected for editing.Example 4: Determination of desired edits to genes selected for editing

[0103] Desired edits are determined for all genes selected for editing in Example 2 or 3. Desired edits are designed to reduce the outlier score of each selected gene in the first cultivar in order to restore expression to be closer to the population consensus. Some desired edits may be designed to replicate an existing allele in another cultivar. Some desired edits may be designed to create a novel allele.

[0104] Some desired edits may be designed to introduce small deletions in promoters to alter expression.

[0105] Some desired edits may be designed to introduce small element insertions to increase expression.

[0106] Some desired edits may be designed to target promoters, enhancers, transcription factor binding sites, splice donor sites, splice acceptor sites, miRNA target loci, siRNA target loci, insulators, introns, 5’ UTRs, or 3’ UTRs.Example 5: Testing of desired edits to genes selected for editing

[0107] Individual edits determined in Example 4 are performed in vitro. The effect of the edit on the RNA expression of the gene is determined by comparing cells with the edit to cells without the edit using methods known in the art, such as qRT-PCR. Edits that do not have the desired effect are exchanged in the list of total desired edits for alternative edits that do have the desired effect.Example 6: Editing a plant

[0108] The desired edits from any of Examples 2, 3, or 4 are introduced into a plant using means known in the art, including, but not limited to, CRISPR / Cas editing, base editing, prime editing, or editing with another site- specific nuclease such as a zinc finger nuclease or TALEN. The edited plant is grown. RNA expression outlier status of the edited genes is reduced, as is confirmed by RNA expression data from the edited plant.Example 7 : Selecting breeding parents based on outlier status

[0109] Outlier scores are calculated as in Example 1 for multiple individuals of a breeding population. Algorithms are used to identify pairs of individuals capable of breeding or hybridizing with one another that minimize the number of loci at which both individualshave high outlier scores in common, and / or to maximize the number of loci wherein the two individuals’ RNA expression levels will compensate for one another. Selected individuals are then bred or hybridized to produce offspring with reduced outlier scores at multiple loci across the genome.

Claims

CLAIMSWhat is claimed is:

1. A method for improving performance of a plant, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of the plant or of a plant population comprising the plant; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in the plant or plant population in comparison to one or more reference transcript expression levels; c. obtaining a distribution of the plurality of outlier scores; d) identifying a subset of the plurality of RNA transcripts with outlier scores beyond a threshold in the distribution; e) determining, for each RNA transcript of the subset of RNA transcripts, a genetic modification that changes the RNA expression of the RNA transcript to have a reduced outlier score magnitude; and f) introducing into the plant the genetic modifications determined for one or more of the RNA transcripts of the subset of RNA transcripts, thereby improving the performance of the plant.

2. The method of claim 1, wherein the method further comprises the step of: g) comparing the performance of the plant comprising the determined genetic modifications to the performance of a control plant.

3. A method of identifying targets for genetic modification to improve performance of a plant, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of the plant or of a plant population comprising the plant; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in the plant or plant population in comparison to one or more reference transcript expression levels; c) obtaining a distribution of the plurality of outlier scores;d) identifying a subset of the plurality of RNA transcripts with outlier scores beyond a threshold in the distribution; and e) determining, for each RNA transcript of the subset of RNA transcripts, a genetic modification that changes the RNA expression of the RNA transcript to have a reduced outlier score magnitude, thereby identifying targets for genetic modification to improve plant performance.

4. The method of claim 3, wherein the method further comprises the step of: f) selecting a subset of targets for genetic modification in a plant from the determined genetic modifications, wherein the selected genetic modifications are predicted to reduce the magnitude of RNA expression outlier scores of the RNA transcripts.

5. A method for selecting two plant individuals to breed or hybridize, comprising: a) providing RNA expression levels for a plurality of RNA transcripts of more than one plant or plant population; b) obtaining a plurality of outlier scores by determining, for each RNA transcript of the plurality of RNA transcripts, based on the RNA expression level, an outlier score for the expression level of the RNA transcript in each of the more than one plants or plant populations in comparison to one or more reference transcript expression levels; c) obtaining a distribution of the plurality of outlier scores for each of the more than one plants or plant populations; d) identifying in each of the more than one plants or plant populations a subset of the plurality of RNA transcripts with outlier scores beyond a specified threshold in the distribution; and e) selecting two plant individuals capable of breeding or hybridizing with one another, wherein the two plant individuals have minimal overlap in their two subsets of RNA transcripts, or wherein the two plant individuals are derived from two plant populations with minimal overlap in their two subsets of RNA transcripts.

6. The method of any one of claims 1-5, wherein the method further comprises steps e’), e”), and e’”) after step e) and before any following steps, comprising: e’) analyzing the effect of one or more genetic modifications determined in step e) on RNA expression;e”) determining one or more alternative genetic modifications for an RNA transcript and performing additional analysis; and e”’) selecting a genetic modification for each RNA transcript in the subset of RNA transcripts.

7. The method of claim 6, wherein the analysis is performed in vitro, in silico, in cell culture, in protoplasts, and / or by Perturb-Seq.

8. The method of claim 6 or claim 7, wherein step e”) is repeated until a genetic modification with a desired outcome is obtained.

9. The method of claim 8, wherein the desired outcome is a reduction of the outlier score for the RNA transcript such that the outlier score is no longer beyond the threshold.

10. The method of any one of claims 1-9, wherein the RNA expression levels are not pooled from multiple individuals.

11. The method of any one of claims 1-9, wherein the RNA expression levels are combined from multiple individuals from the plant population.

12. The method of any one of claims 1-11, wherein the outlier score is determined using RNA expression levels across multiple tissues.

13. The method of claim 12, wherein the RNA expression levels are from a sample comprising multiple tissues, or are pooled from multiple tissues.

14. The method of any one of claims 1-13, wherein the outlier scores are determined using RNA expression levels from a single tissue.

15. The method of any one of claims 1-14, wherein the outlier score is determined using RNA expression levels across multiple members of a gene family.

16. The method of any one of claims 1-14, wherein the outlier score is determined using RNA expression levels across multiple members of a gene interaction network.

17. The method of any one of claims 1-16, wherein the outlier score is determined using protein expression levels in addition to RNA expression levels.

18. The method of any one of claims 1-17, wherein the outlier score is determined using splice isoform ratios.

19. The method of any one of claims 1-18, wherein the RNA expression levels are from whole genome RNA-seq data.

20. The method of any one of claims 1-19, wherein the RNA expression levels are derived from a micro level tissue.

21. The method of claim 20, wherein the micro level tissue is selected from the group consisting of shoot apical meristem, inflorescence meristem, immature male embryos, immature female embryos, developing anthers, and root hairs.

22. The method of any one of claims 20 or 21, wherein the RNA expression levels are obtained by single cell sequencing.

23. The method of any one of claims 20 or 21, wherein the micro level tissue is obtained by laser capture microdissection.

24. The method of any one of claims 1-23, wherein the RNA expression levels are derived from tissue isolated at a specific plastochron.

25. The method of any one of claims 1-19, wherein the RNA expression levels are derived from macro level tissue.

26. The method of claim 25, wherein the macro level tissue is selected from the group consisting of leaf, root, internode, anther, tassel, ear shank, kernel, and ligule.

27. The method of any one of claims 1-19, wherein the RNA expression levels are derived from protoplasts, from undifferentiated cells, from plant calli, or from plant cell suspension cultures.

28. The method of any one of claims 1-27, wherein the plurality of RNA transcripts comprises mRNAs.

29. The method of any one of claims 1-28, wherein the plurality of RNA transcripts comprises IncRNAs.

30. The method of any one of claims 1-29, wherein the plurality of RNA transcripts comprises miRNAs.

31. The method of any one of claims 1 -30, wherein the plurality of RNA transcripts comprises siRNAs.

32. The method of any one of claims 1-31, wherein the plurality of RNA transcripts comprises transcripts associated with flowering, transcripts associated with growth, transcripts associated with biotic stress resistance, and / or transcripts associated with abiotic stress resistance.

33. The method of any one of claims 1-32, wherein the plurality of RNA transcripts consists of all RNA transcripts.

34. The method of any one of claims 1-33, wherein the reference transcript expression level is the RNA expression level of the RNA transcript in another plant of the same species.

35. The method of any one of claims 1-33, wherein the reference transcript expression level is the consensus transcript level within a population of plants of the same species.

36. The method of any one of claims 1-33, wherein the reference transcript expression level is the consensus transcript level across multiple cultivars of the same species.

37. The method of any one of claims 1-33, wherein the reference transcript expression level is the consensus transcript level of orthologs across multiple species.

38. The method of any one of claims 1-37, wherein the outlier score is expressed as the deviation of the mean scaled by the standard deviation of the population.

39. The method of any one of claims 1-37, wherein the outlier score is a Z-score.

40. The method of any one of claims 1-37, wherein the outlier score is a burden score tied to a Z-score.

41. The method of any one of claims 39 or 40, wherein the Z-score is a non-parametric Z- score.

42. The method of any one of claims 39 or 40, wherein the Z-score is derived from provided estimates of the dispersal or from the defined point mass of the distribution.

43. The method of any one of claims 1-37, wherein the outlier score is determined by autoencoder-based outlier detection, wherein the autoencoder-based outlier detection comprises an Encoder and a Decoder, wherein the Encoder and the Decoder each comprise one or more non-linear transformations.

44. The method of claim 43, wherein the autoencoder takes an input, wherein the input to the autoencoder comprises RNA expression measurements.

45. The method of claim 43 or claim 44, wherein the Encoder of the autoencoder comprises an artificial neural network (ANN).

46. The method of any one of claims 43-45, wherein the Decoder of the autoencoder comprises an artificial neural network (ANN).

47. The method of any one of claims 43-46, wherein the outlier score is determined by anomaly detection with attributed networks.

48. The method of claim 47, wherein the outlier score is determined for a set of genes in a known gene regulatory network.

49. The method of any one of claims 47 or 48, wherein the autoencoder takes an input, wherein the input to the autoencoder comprises a vector comprising measured expression values for a single gene across multiple tissues.

50. The method of any one of claims 47-49, wherein the Encoder is constructed as a graph neural network (GNN).

51. The method of any one of claims 47-50, wherein the Decoder is constructed as a graph neural network (GNN).

52. The method of any one of claims 43-51, wherein the non-linear transformations of the Encoder, the Decoder, or both the Encoder and Decoder are modified by tuning a set of parameters (0).

53. The method of claim 52, wherein the parameters (0) are trained by passing an input vector through the model to produce an output vector, and iteratively tuning the parameters (0) via backpropagation through the series of non-linear transformations by minimizing a differentiable reconstruction error function between the output vector and the input vector.

54. The method of claim 53, wherein the differentiable reconstruction error function is the mean squared error (MSE).

55. The method of any one of claims 1-37, wherein the outlier score is determined by multivariate outlier detection methods, wherein information is borrowed across genes in a local network and across tissues.

56. The method of claim 55, wherein the multivariate outlier detection methods comprise use of the Mahalanobis distance.

57. The method of any one of claims 1-56, wherein the threshold is a one-tailed threshold on the distribution, a two-tailed threshold on the distribution, a p-value, or a set number of genes at one or more tails of the distribution.

58. The method of claim 57, wherein the threshold is set at 50 genes at the upper tail of the distribution.

59. The method of any one of claims 1-58, wherein the genetic modification determined for an individual RNA transcript creates a novel allele.

60. The method of any one of claims 1-58, wherein the genetic modification determined for an individual RNA transcript replicates an alternative allele not present in the plant.

61. The method of claim 60, wherein the alternative allele is an allele present in the plant from which the reference transcript expression level is derived.

62. The method of any one of claims 60 or 61, wherein the alternative allele is an allele with a reduced outlier score from another plant.

63. The method of any one of claims 1-62, wherein the genetic modification determined for an individual RNA transcript targets a DNA sequence element selected from the group consisting of a promoter, an enhancer, a transcription factor binding site, a splice donor site, a splice acceptor site, an miRNA binding site, an siRNA binding site, an insulator, an intron, an exon, a 5’ UTR, and a 3’ UTR.

64. The method of claim 63, wherein the genetic modification determined for an individual RNA transcript is a small deletion in the promoter.

65. The method of claim 63, wherein the genetic modification determined for an individual RNA transcript is a small element insertion to increase expression.

66. The method of any one of claims 1-65, wherein the genome of the plant is modified using a nuclease selected from the group consisting of a Cas nuclease, a TALE nuclease, a zinc finger nuclease, a meganuclease, or an engineered nuclease.

67. The method of claim 66, wherein the Cas nuclease is a Cas9 or Cas 12 nuclease.

68. The method of any one of claims 1-67, wherein the reduced outlier score magnitude is no longer beyond the threshold, and / or is reduced by at least one standard deviation of the distribution.

69. The method of any one of claims 1-68, wherein the plant is selected from the group consisting of corn, soy, and wheat.

70. The method of any one of claims 1-69, wherein the RNA expression levels are obtained from a plant from an inbred panel, a hybrid panel, or an asexually-propagated population.

71. The method of any one of claims 1-70, wherein improved performance of the plant is measured by a metric selected from the group consisting of growth, yield, survival, biomass, number of kernels, ear weight, and nutritional content.

72. The method of any one of claims 5-71, wherein:(a) one or more RNA transcripts present in the subset of RNA transcripts of the first of the twoselected plant individuals are not present in the subset of RNA transcripts of the second of the two selected plant individuals;(b) between two and twenty RNA transcripts present in the subset of RNA transcripts of the first of the two selected plant individuals are not present in the subset of RNA transcripts of the second of the two selected plant individuals;(c) the selected two plants capable of breeding or hybridizing with one another have less than 50% of the subset of RNA transcripts overlapping between their two subsets of RNA transcripts; and / or(d) the selected two plants capable of breeding or hybridizing with one another have no overlap in their two subsets of RNA transcripts.

73. A plant produced by the method of any one of claims 1, 2, and 5-72.

74. A seed produced by the method of any one of claims 5-72, or by the plant of claim 73.

75. A plant grown from the seed of claim 74.