Genomic prediction of discordant gene-phenotype relationships between species

Abstract

Studies of non-human species are often pursued to validate or extend the biological insights of studies exploring gene-phenotype relationships in humans. When the results of non-human studies do not align with those of human studies, various questions arise regarding the study design, contextual factors, reagents and assays employed, and other related aspects. It is arguable that outside of the experimental conditions, there could be biological reasons why non-human and human studies produce discordant results exploring gene-phenotype relationships. These reasons might include the possibility that the gene in question has evolved a different function in one of the species or that the regulatory mechanisms governing its expression have undergone changes. This study assessed the potential of certain gene features, such as sequence conservation levels and aspects of regulatory architecture, to explain the concordant and discordant results obtained from studies involving humans and mice.

Description

GENOMIC PREDICTION OF DISCORDANT GENE-PHENOTYPERELATIONSHIPS BETWEEN SPECIESRELATED APPLICATIONS

[0001] This application claims the benefit of U.S. provisional patent application 63 / 592,554, filed October 23, 2023, to The Translational Genomics Research Institute, titled “GENOMIC PREDICTION OF DISCORDANT GENE-PHENOTYPE RELATIONSHIPS BETWEEN SPECIES,” the entirety of the disclosure of which is hereby incorporated by this reference.STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0002] This invention was made with government support under grant numbers AG064706, AG023122, and AG065169 awarded by the National Institutes of Health. The government has certain rights in the invention.TECHNICAL FIELD

[0003] This document relates to the field of comparative genomics combined with computational and statistical analysis approaches to predict whether a phenotype is concordant between two species.BACKGROUND

[0004] Analysis of concordant and discordant gene-phenotype relationships between two species, such as humans and mice, involves the comparison of the genomic features of both species to determine why certain genes that are associated with a particular phenotype in one species may not be associated with the same phenotype in the other species. There is a need in the art for improved ability to determine how genomic features contribute to discordant gene-phenotype associates.SUMMARY

[0005] In one aspect, the present invention provides for a method of predicting concordance and discordance in gene-phenotype relationships between a first species and a second species, the method comprising: selecting one or more genes associated with a phenotype, wherein the selected genes are associated with the phenotype (a) in both the first species and the second species, (b) in the first species only, and / or (c) in the second species only; determining a pair-wise correlation of a plurality of genomic predictors for the one or more selected genes in the first species compared to the second species; and predicting the concordance or discordance for the phenotype between the first and second species based on the genomic predictors using a statistical analysis for individual genes. In certain embodiments, the plurality of genomic predictors comprise: a pair- wise ortholog score, whereby the conservation level of each gene between the species is determined, syntenic block harboring of each gene between the species, the number of paralogs of each gene in each species, the number of regulatory elements associated with each gene in each species, number of tissues each gene is expressed in for each species, whereby expression patterns of each gene is determined, the connectivity, centrality, or position in a protein-protein interaction (PPI) network for each gene in each species, and / or the number of functional processes identified for each gene. In certain embodiments, the first species is human and the second species is a nonhuman animal. In certain embodiments, the statistical analysis is a pair-wise t-test. In certain embodiments, the method further comprises calculating the probability of the concordance or discordance based on the genomic predictors for each gene using a multi-predictor linear regression model.

[0006] In certain embodiments, the phenotype is longevity. In certain embodiments, the phenotype is a disease phenotype. In certain embodiments, the one or more genes are selected from the GenAge database. In certain embodiments, the one or more genes are selected from genes linked to the phenotype based on genome-wide association studies (GWAS) for the phenotype. In certain embodiments, the one or more selected genes are all expressed in the same tissue or cell type.

[0007] In certain embodiments, the method further comprises removing from the analysis any genomic predictors for which a threshold amount of data is missing from the database. In certain embodiments, the method additionally comprises removing from analysis the genomic predictorshaving a threshold amount of data mi ssing for the one or more genes to ensure that there are enough genomic predictors for each gene to be confident in the predictions about concordant or discordant effects of the phenotype between the species. In certain embodiments, the threshold amount is more than 40% of data being missing.

[0008] In certain embodiments, the pair-wise correlation comprises a Pearson Correlation coefficient to determine correlation between the genomic predictors between the first species and the second species. In certain embodiments, dimension reduction is used to analyze the pair-wise correlations to identify combinations of the genomic predictors that have the most predictive power. In certain embodiments, the dimension reduction comprises a principal component analysis (PC A), whereby linear combinations of the genomic predictors that have the most predictive power are identified. In certain embodiments, a linear regression model is used in the concordance or discordance analysis. In certain embodiments, predicted probability is used to evaluate the accuracy of the linear regression model. In certain embodiments, a multivariate distance matrix regression (MDMR) method is used to analyze the relationship between multiple genomic predictors and the one or more genes. In certain embodiments, the method further comprises using the MDMR method to analyze different random sets of genes from a group of genes present in neither the first species nor second species to check for consistency of analysis.

[0009] In certain embodiments, gene set based enrichment analysis (GSEA) is used to identify functional processes in the concordance or discordance analysis. In certain embodiments, a tau index is used in the concordance or discordance analysis, wherein the tau index indicates how specific or broadly each gene is expressed across tissues.

[0010] In certain embodiments, the method further comprises selecting the second species as a non-human animal model to test a medicament for modulating the phenotype in the first species if concordance for the phenotype between the first and second species is predicted.

[0011] In certain embodiments, the one or more genes associated with a phenotype are selected by determining genes associated with the phenotype in the first and second species; and selecting genes (a) present in both the first species and the second species or (b) present in the first species only.

[0012] In another aspect, the present invention provides for a method of identifying an agent or target gene for modulating a phenotype in a subject comprising contacting a non-human animal model with one or more agents specific for targeting one or more genes, wherein the phenotypemodulated in the non-human animal model is concordant with the phenotype in a human subject, wherein concordance of the phenotype is determined by a method comprising: selecting one or more genes associated with a phenotype, wherein the selected genes are associated with the phenotype (a) in both the first species and the second species, (b) in the first species only, and / or (c) in the second species only; determining a pair-wise correlation of a plurality of genomic predictors for the one or more selected genes in the first species compared to the second species; and predicting the concordance or discordance for the phenotype between the first and second species based on the genomic predictors using a statistical analysis for individual genes. In certain embodiments, the plurality of genomic predictors comprise: a pair-wise ortholog score, whereby the conservation level of each gene between the species is determined, syntenic block harboring of each gene between the species, the number of paralogs of each gene in each species, the number of regulatory elements associated with each gene in each species, number of tissues each gene is expressed in for each species, whereby expression patterns of each gene is determined, the connectivity, centrality, or position in a protein-protein interaction (PPI) network for each gene in each species, and / or the number of functional processes identified for each gene. In certain embodiments, the first species is human and the second species is a non-human animal. In certain embodiments, the statistical analysis is a pair-wise t-test. In certain embodiments, the method further comprises calculating the probability of the concordance or discordance based on the genomic predictors for each gene using a multi-predictor linear regression model.

[0013] In certain embodiments, the phenotype is longevity. In certain embodiments, the phenotype is a disease phenotype. In certain embodiments, the one or more genes are selected from the GenAge database. In certain embodiments, the one or more genes are selected from genes linked to the phenotype based on genome-wide association studies (GWAS) for the phenotype. In certain embodiments, the one or more selected genes are all expressed in the same tissue or cell type.

[0014] In certain embodiments, the method further comprises removing from the analysis any genomic predictors for which a threshold amount of data is missing from the database. In certain embodiments, the method additionally comprises removing from analysis the genomic predictors having a threshold amount of data missing for the one or more genes to ensure that there are enough genomic predictors for each gene to be confident in the predictions about concordant or discordanteffects of the phenotype between the species. In certain embodiments, the threshold amount is more than 40% of data being missing.

[0015] In certain embodiments, the pair-wise correlation comprises a Pearson Correlation coefficient to determine correlation between the genomic predictors between the first species and the second species. In certain embodiments, dimension reduction is used to analyze the pair-wise correlations to identify combinations of the genomic predictors that have the most predictive power. In certain embodiments, the dimension reduction comprises a principal component analysis (PC A), whereby linear combinations of the genomic predictors that have the most predictive power are identified. In certain embodiments, a linear regression model is used in the concordance or discordance analysis. In certain embodiments, predicted probability is used to evaluate the accuracy of the linear regression model. In certain embodiments, a multivariate distance matrix regression (MDMR) method is used to analyze the relationship between multiple genomic predictors and the one or more genes. In certain embodiments, the method further comprises using the MDMR method to analyze different random sets of genes from a group of genes present in neither the first species nor second species to check for consistency of analysis.

[0016] In certain embodiments, gene set based enrichment analysis (GSEA) is used to identify functional processes in the concordance or discordance analysis. In certain embodiments, a tau index is used in the concordance or discordance analysis, wherein the tau index indicates how specific or broadly each gene is expressed across tissues.

[0017] In certain embodiments, the one or more genes associated with a phenotype are selected by determining genes associated with the phenotype in the first and second species; and selecting genes (a) present in both the first species and the second species or (b) present in the first species only.

[0018] The foregoing and other aspects, features, and advantages will be apparent from the DESCRIPTION and DRAWINGS, and from the CLAIMS if any are included.BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Implementations will hereinafter be described in conjunction with the appended and / or included DRAWINGS, where like designations denote like elements, and:

[0020] FIG. 1 presents a table showing genomic feature characterization.

[0021] FIGS. 2A and 2B illustrate regression analysis to investigate relationships among predictor variables for variations between two groups.

[0022] FIG. 3 presents multivariate distance matrix regression (MDMR) among the predictor variables.

[0023] FIG. 4 shows overlap between human and mouse genes known to be associated with longevity.

[0024] FIG. 5 illustrates genomic feature characterization. Genomic features refer to the distinctive characteristics or components of a genome (functional elements, regulatory elements, functional connections, epigenetic signature etc.). Understanding the various genomic features and how they interact with each other is critical for understanding the function of genes and the development of diseases.

[0025] FIGS. 6A-6E show pair-wise relationship between each pair of variables in the datasets.

[0026] FIGS. 7A-7C present principal component analysis (PCA) among the predictor variables.

[0027] FIGS. 8 and 9 show multivariate distance matrix regression (MDMR) among the predictor variables.

[0028] FIG. 10 illustrates comparative functional enrichment analysis.

[0029] FIG. 11 cross-species aging genes comparison- human and non-human species.

[0030] FIGS. 12A-12C show cross-species aging genes comparison- strength of the association in humans and the other non-human species.

[0031] FIGS. 13A-13D show cross-species gene comparison — gene-set comparison.

[0032] FIGS. 14A-14H show pair-wise correlation of predictors.

[0033] FIGS. 15A-15B illustrate regression models to contrast genes between groups.

[0034] FIGS. 16A-16D illustrate cross-species aging gene comparison — principal component analysis.

[0035] FIGS. 17A-17D show human tissue specific gene expression comparison within the groups using a Tau index.

[0036] Accordingly, it is to be understood that the embodiments of the invention herein described are merely illustrative of the application of the principles of the invention. Referenceherein to details of the illustrated embodiments is not intended to limit the scope of the claims, which themselves recite those features regarded as essential to the invention.DETAILED DESCRIPTION

[0037] Detailed aspects and applications of the disclosure are described below in the following drawings and detailed description of the technology. Unless specifically noted, it is intended that the words and phrases in the specification and the claims be given their plain, ordinary, and accustomed meaning to those of ordinary skill in the applicable arts.

[0038] In the following description, and for the purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the various aspects of the disclosure. It will be understood, however, by those skilled in the relevant arts, that embodiments of the technology disclosed herein may be practiced without these specific details. It should be noted that there are many different and alternative configurations, devices and technologies to which the disclosed technologies may be applied. The full scope of the technology disclosed herein is not limited to the examples that are described below.

[0039] The singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to “a step” includes reference to one or more of such steps. Reference to a particular numerical value includes at least that particular value, unless the context clearly dictates otherwise.

[0040] The word “exemplary,” “example,” or various forms thereof are used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as “exemplary” or as an “example” is not necessarily to be construed as preferred or advantageous over other aspects or designs. Furthermore, examples are provided solely for purposes of clarity and understanding and are not meant to limit or restrict the disclosed subject matter or relevant portions of this disclosure in any manner. It is to be appreciated that a myriad of additional or alternate examples of varying scope could have been presented but have been omitted for purposes of brevity.

[0041] When a range of values is expressed, another embodiment includes from the one particular value and / or to the other particular value. Similarly, when values are expressed asapproximations, by use of the antecedent “about,” it will be understood that the particular value forms another embodiment. All ranges are inclusive and combinable.

[0042] The term “plurality”, as used herein, means more than one.

[0043] Throughout the description and claims of this specification, the words “comprise” and “contain” and variations of the words, for example “comprising” and “comprises”, mean “including but not limited to”, and are not intended to (and do not) exclude other components.

[0044] As required, detailed embodiments of the present disclosure are included herein. It is to be understood that the disclosed embodiments are merely exemplary of the invention that may be embodied in various forms. Therefore, specific structural and functional details disclosed herein are not to be interpreted as limits, but merely as a basis for teaching one skilled in the art to employ the present invention. The specific examples below will enable the disclosure to be better understood. However, they are given merely by way of guidance and do not imply any limitation.

[0045] The present disclosure may be understood more readily by reference to the following detailed description taken in connection with the accompanying figures and examples, which form a part of this disclosure. It is to be understood that this disclosure is not limited to the specific materials, devices, methods, applications, conditions, or parameters described and / or shown herein, and that the terminology used herein is for the purpose of describing particular embodiments by way of example only and is not intended to be limiting of the claimed inventions. The term “plurality”, as used herein, means more than one. When a range of values is expressed, another embodiment includes from the one particular value and / or to the other particular value. Similarly, when values are expressed as approximations, by use of the antecedent “about,” it will be understood that the particular value forms another embodiment. All ranges are inclusive and combinable.

[0046] More specifically, this disclosure, its aspects and embodiments, are not limited to the specific material types, components, methods, or other examples disclosed herein. Many additional material types, components, methods, and procedures known in the art are contemplated for use with particular implementations from this disclosure. Accordingly, for example, although particular implementations are disclosed, such implementations and implementing components may comprise any components, models, types, materials, versions, quantities, and / or the like as is known in the art for such systems and implementing components, consistent with the intended operation.

[0047] The present disclosure provides for methods of predicting concordance and discordance in gene-phenotype relationships between a first species and a second species by determining correlations of genomic and epigenetic features between the two species for a selected set of genes associated with a phenotype of interest, such that the second species can be used as a model organism for identifying treatments for the first species.1. Overview

[0048] To determine how genomic features contribute to concordant and discordant genephenotype associations, Applicants used comparative genomics together with a number of computational and statistical analysis approaches. One of the uses of such an approach would be to improve selection of a non-human animal species to use in early-stage pharmaceutical testing. Preclinical studies typically involve non-human animal studies that serve as models to study disease treatment of a particular pharmaceutical product. The better the non-human animal models predict the interaction of the pharmaceutical product with human patients, the more efficient drug development will be. If the non-human animal model does not correlate with human patients, potential pharmaceutical candidates may be discarded unnecessarily during early testing.

[0049] In some embodiments, genes linked to human and mouse longevity from a curated database are examined. Other embodiments use a larger data set with a plurality of species.

[0050] In certain embodiments, Applicants used the methods more widely by incorporating genome-wide association studies (GWAS) findings on human longevity and differentially expressed genes linked to the maximum lifespan in mice.

[0051] Studies of non-human species are often pursued to validate or extend the biological insights of studies exploring gene-phenotype relationships in humans. Questions about the design of the studies, the contexts within which they are pursued, the reagents and assays used, etc. arise when the results of the non-human studies do not match those of the human studies. It is arguable that outside of the experimental conditions, there could be biological reasons why non-human and human studies produce discordant results exploring gene-phenotype relationships. For example, the gene in question could have evolved a different function in one of the species, or the way it is regulated has changed. Applicants explore the ability of features of genes found to show concordant and discordant results in studies involving humans and mice. These features include sequence conservation levels of genes and aspects of their regulatory architecture.

[0052] Suppose one can manipulate a gene and its expression (e.g., by providing a drug known target the gene's activity, knocking the gene out with genetic engineering techniques, or inducing or identifying mutations in the gene that cause it to behave in certain ways). If this gene contributes to a phenotype, like heart disease or diabetes, then by modifying the gene, heart disease or diabetes could be prevented or treated. If one uses the drug or genetic engineering technique in a mouse and sees an effect, then that would mean the drug or genetic engineering did what it was supposed to do. If it does the same thing in humans, then the argument is that the mouse is a good model for humans. However, if the drug or genetic engineering does not do anything in the mouse, but does in humans (or vice versa) then the mouse is not a good model for humans. This kind of discordance in effect is seen quite often when, e.g., drugs are tested. If one could predict when such discordance in effect occurs based on the features or properties of the genes to be targeted or modulated, then one could avoid using species that are not good models for humans. Applicants discovered that one can use various characteristics of the gene in non-human and human genomes (DNA sequence, expression patterns, etc.) to predict when discordance and concordance are likely. These predictions can point to which species make sense to use as models for humans and also explain why some drugs or interventions gave discordance effects in the two species.

[0053] Applicants have developed predictive models, using various data analytic techniques, of concordance and discordance of gene-phenotype relationships between species that could be used to prioritize research efforts or put into context specific findings of research involving non- human species. This could have implications for drug development and biomarker evaluation.2. Terminology and Definitions

[0054] In describing the embodiments herein, the following terminology will be used in accordance with the definitions and explanations set out below. Notwithstanding, other terminology, definitions, and explanations may be found throughout this document as well.

[0055] As used herein, “concordance” is a term used in its broadest sense and may refer to the amount of similarity in phenotype between a set of individuals. Concordance as used in genetics means the probability that a pair of individuals will both have a certain characteristic, given that one of the pair has the characteristic. In contrast, “discordance” occurs when a similar trait is not shared by the pair.

[0056] As used herein “correlation” is a term used in its broadest sense and may refer to observable molecular characteristics that increase or decrease with a specific condition (e.g., between species). As used herein “correlation analysis” refers to a statistical method used to measure the strength of the linear relationship between two variables and compute their association. In other words, correlation analysis calculates the level of change in one variable due to the change in the other. Negative or inverse correlation describes when two variables tend to move in opposite size and direction from one another, such that when one increases the other variable decreases, and vice- versa. A positive correlation exists when one variable tends to decrease as the other variable decreases, or one variable tends to increase when the other increases.

[0057] As used herein, “Protein-Protein Interaction” or “PPI” networks is a term used in its broadest sense and may refer to any networks of interacting proteins. Networks of interacting proteins in diseases’ pathways have been analyzed for identifying candidate drug targets. PPI networks for different species can be obtained using a database. For example, STRING is a database of known and predicted protein-protein interactions (see, e.g., Szklarczyk D, Kirsch R, Koutrouli M, et al. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res. 2023;51(Dl):D638-D646). The STRING database currently covers 59,309,604 proteins from 12,535 organisms.

[0058] As used herein, a “genomic predictor” is any genomic or epigenomic or functional variable that can be compared between two species. Genomic predictor includes, but is not limited to differences in nucleotide sequence, protein sequence, regulatory sequences, gene locus, chromatin looping, gene expression, tissue or cell type specific differences, interactions between genes and gene products (e.g., proteins), transcription factor binding, protein modifications associated with a gene or gene locus, functional process associated with genes, and related genes (e.g., paralogs).

[0059] As used herein, “syntenic block” or “synteny” refers the conservation of blocks of order within two sets of chromosomes that are being compared with each other. In classical genetics, syntenic genes were originally defined as genes that lie on the same chromosome. As used herein, a synteny block can also be defined as a region of genome sequence spanning a number of genes that are orthologous and co-arranged with another genome, where orientation is not considered(see, e g., Liu D, Hunt M, Tsai IT Inferring synteny between genome assemblies: a systematic evaluation. BMC Bioinformatics. 2018; 19(1):26).

[0060] As used herein, “orthologs” refers to genes in different species that evolved from a common ancestral gene by speciation.

[0061] As used herein, “paralogs” refers to a set of homologous genes that have diverged from each other as a consequence of genetic duplication or a pair of genes that derives from the same ancestral gene and now reside at different locations within the same genome.

[0062] As used herein, the term “regulatory element” is intended to include transcription factor binding sites, promoters, enhancers, internal ribosomal entry sites (IRES), and other expression control elements (e.g., transcription termination signals, such as polyadenylation signals and poly- U sequences). A transcription factor binding site or a specific regulatory sequence can be a sequence that is in contact with a promoter, a sequence within an enhancer, or an activator binding site (e.g., in contact by looping or 3D architecture of a genome). A tissue-specific promoter may direct expression primarily in a desired tissue of interest, such as muscle, neuron, bone, skin, blood, specific organs (e.g., liver, pancreas), or particular cell types (e.g., lymphocytes). Regulatory elements may also direct expression in a temporal-dependent manner, such as in a cell-cycle dependent or developmental stage-dependent manner, which may or may not also be tissue or celltype specific.

[0063] As used herein, the term “functional processes” refers to functions associated with each gene. The functions can include biological process, molecular function, and cellular component. Biological process refers to a biological objective to which the gene or gene product contributes. Molecular function is defined as the biochemical activity (including specific binding to ligands or structures) of a gene product. Cellular component refers to the place in the cell where a gene product is active. In example embodiments, “functional processes” refers to gene ontology (GO), which focuses on the function of the genes and gene products and provides the logical structure of the biological functions ('GO terms'). The Gene Ontology (GO) knowledgebase is the world’s largest source of information on the functions of genes (see, e.g., Ashbumer M, Ball CA, Blake JA, et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet. 2000;25(l):25-29; and Gene Ontology Consortium, Aleksander SA, Balhoff J, et al. The Gene Ontology knowledgebase in 2023. Genetics. 2023;224(l):iyad031).

[0064] As used herein, “genome-wide association studies (GWAS)” can refer to the results of testing hundreds of thousands of genetic variants across many genomes to find those statistically associated with a specific trait or disease (see, e.g., Uffelmann, E., Huang, Q.Q., Munung, N.S. et al. Genome-wide association studies. Nat Rev Methods Primers 2021;l, 59). Genome-wide association studies (GWAS) aim to identify associations of genotypes with phenotypes by testing for differences in the allele frequency of genetic variants between individuals who are ancestrally similar but differ phenotypically. GWAS can consider copy-number variants or sequence variations in the human genome, although the most commonly studied genetic variants in GWAS are single-nucleotide polymorphisms (SNPs). GWAS typically report blocks of correlated SNPs that all show a statistically significant association with the trait of interest, known as genomic risk loci.

[0065] As used herein, a “machine learning algorithm” is a set of rules or processes used by an artificial intelligence (Al) system to conduct tasks, most often to discover new data insights and patterns, or to predict output values from a given set of input variables. Algorithms enable machine learning (ML) to learn. Machine learning algorithms are computational models that allow computers to understand patterns and forecast or make judgments based on data without explicit programming. There are four types of machine learning algorithms: 1. supervised learning, 2. unsupervised learning, 3. reinforcement learning, and 4. ensemble learning. Supervised learning includes classification, which includes logistic regression, support vector machines (svm), k- nearest neighbors (k-NN), naive bayes, decision trees, random forest, gradient boosting (e.g., XGBoost, LightGBM, CatBoost), and neural networks (e g., multilayer perceptron). Supervised learning also includes regression, which includes linear regression, ridge regression, lasso regression, support vector regression (SVR), decision trees regression, random forest regression, gradient boosting regression, and neural networks regression. Unsupervised learning includes clustering, which includes k-means, hierarchical clustering, DBSCAN (Density-Based Spatial Clustering of Applications with Noise), and Gaussian Mixture Models (GMM). Unsupervised learning also includes dimensionality reduction, which includes principal component analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE), linear discriminant analysis (LDA), independent component analysis (ICA), and UMAP (Uniform Manifold Approximation and Projection). Unsupervised learning also includes association, which includes apriori algorithm and eclat algorithm. Reinforcement learning includes model-free methods, which includes Q-Leaming,deep Q-network (DQN), SARSA (State-Action-Reward-State-Action), and policy gradient methods (e.g., REINFORCE). Reinforcement learning also includes model-based methods, which includes deep deterministic policy gradient (DDPG), proximal policy optimization (PPO), and trust region policy optimization (TRPO). Reinforcement learning also includes value-based methods, which includes monte carlo methods and temporal difference (TD) learning. Ensemble learning includes bagging (e.g., random forest), boosting (e.g., AdaBoost, gradient boosting), and stacking.

[0066] As used herein, the term “linear regression” refers to modeling the relationship between dependent and independent variables using a linear approach.

[0067] As used herein, the term “dimension reduction” refers to the transformation of data from a high-dimensional space into a low-dimensional space so that the low-dimensional representation retains some meaningful properties of the original data, ideally close to its intrinsic dimension. Methods are commonly divided into linear and nonlinear approaches. Approaches can also be divided into feature selection and feature extraction. Dimensionality reduction can be used for noise reduction, data visualization, cluster analysis, or as an intermediate step to facilitate other analyses. The main linear technique for dimensionality reduction, principal component analysis (PCA), performs a linear mapping of the data to a lower-dimensional space in such a way that the variance of the data in the low-dimensional representation is maximized (see, e.g., Lever, J, Krzywinski, M, Altman, N. Principal component analysis. Nat Methods 2017; 14, 641-642). In practice, the covariance (and sometimes the correlation) matrix of the data is constructed and the eigenvectors on this matrix are computed. The eigenvectors that correspond to the largest eigenvalues (the principal components) can now be used to reconstruct a large fraction of the variance of the original data. PCA reduces the dimensionality of data by transforming it to a new set of orthogonal features (principal components) that capture the maximum variance.

[0068] Any reference to a gene is a reference made to the gene and the gene product (e.g., protein).3. Methods of Predicting Concordance Between Two Species

[0069] In an aspect, genomic predictors are compared to determine the degree of correlation between two species. The genomic predictors are analyzed for a set of genes associated with a phenotype of interest. In example embodiments, the association of a gene with the phenotype ofinterest is known in only one of the species, however, the gene may be associated with the phenotype in both species or only the one species. Thus, genes that are associated with the phenotype in either species or both species can be selected. The phenotype of interest can be any known or newly discovered phenotype. Non-limiting phenotypes include longevity and disease phenotypes. For example, genes may be associated with a longer or shorter lifespan or genes may be associated with an increased risk for a disease (e.g., heart disease, cancer, obesity). In example embodiments, machine learning algorithms can be used to analyze the correlation of the genomic predictors for the set of genes between the two species. In example embodiments, a treatment targeting a phenotype or an agent for modulating a phenotype in humans can be identified using a non-human animal model if there is concordance between the two species for the phenotype.

[0070] In example embodiments, different species can be selected from any two species. In preferred embodiments, one species is a human and the second species is a eukaryotic species, more preferably, a non-human animal, more preferably, a mammal, such as a rodent or primate. Non-limiting examples of species include primates (e.g., chimpanzees and other apes and monkey species), farm animals (e.g., cattle, sheep, pigs, goats and horses), domestic mammals (e.g., dogs and cats), laboratory animals (e.g., rodents such as mice, rats, and guinea pigs), and birds (e.g., domestic, wild and game birds such as chickens, turkeys and other gallinaceous birds, ducks, geese, and the like). In some embodiments, the species is a mammal. In further embodiments, the species is a human.Selecting Genes Associated with a Phenotype

[0071] In example embodiments, one or more genes associated with a phenotype of interest are selected to predict concordance between a first and second species, whereby testing medicaments in a non-human animal model that is concordant can be reliably used to identify treatments for a disease in humans. As used herein, a phenotype of interest refers to a phenotype that is present in a first species (e.g., human) that is to be studied in a second species. In example embodiments, 1, 10, 20, 40, 50, 100, 150, 1000, or more genes are selected. In example embodiments, genes associated with a phenotype in at least one of the species is selected. In example embodiments, the phenotypes can be any phenotype related to health or disease. For example, non-limiting phenotypes include longevity, cancer, heart disease, obesity, high blood pressure, diabetes, inflammatory bowel disease (IBD), autoimmune diseases, Parkinson’s disease, multiple sclerosis, psychiatric phenotypes (e.g., schizophrenia, autism spectrum disorders,attention deficit hyperactivity disorder, major depressive disorder, and bipolar disorder), or any phenotype present in a GWAS dataset (e.g., The UK Biobank (UKBBK)). In example embodiments, genes known to be associated with a phenotype of interest are selected. The genes may be part of a pathway identified as being associated with a phenotype of interest. Genes may be shown to interact with genes associated with a phenotype of interest. In example embodiments, differentially expressed genes between subjects having the phenotype and subjects not having the phenotype are selected. For example, differentially expressed genes between subjects with heart disease and subjects without heart disease are selected.Selecting Genes Associated to a Phenotype Using GWAS

[0072] In example embodiments, GWAS is used to identify genes that are associated with a phenotype of interest. For example, GWAS provides genetic variants (e.g., SNPs, amplifications, deletions, insertions) associated with a phenotype of interest and these genetic variants can be linked to genes (i.e., variants are derived from the GWAS). As used herein “genetic variants” or “variants” refers to any difference in DNA among individuals. Genetic variation is caused by variation in the order of bases in the nucleotides in genomic loci. Examination of DNA has shown genetic variation in both coding regions and in the non-coding intron region of genes. Genetic variations may be present in regulatory regions (e.g., promoters, enhancers, repressors) or nonprotein coding genes (e.g., IncRNA, miRNA, snRNA). In example embodiments, the genetic variants are single-nucleotide polymorphisms (SNPs). A SNP is a substitution of a single nucleotide that occurs at a specific position in the genome, where each variation is present to some appreciable degree within a population (e.g. >1%). In certain embodiments, genetic variants associated with the phenotype of interest are identified using a biobank or database (see, e.g., UK Biobank; By croft et al., The UK Biobank resource with deep phenotyping and genomic data. Nature 562, 203-209 (2018); 1000 Genomes Project Consortium. A global reference for human genetic variation. Molecular cell, 526(7571):68-74, 2015; and Sollis E, Mosaku A, Abid A, et al. The NHGRI-EBI GWAS Catalog: knowledgebase and deposition resource. Nucleic Acids Res. 2023;51(Dl):D977-D985).

[0073] In example embodiments, variants that are a certain distance from a gene are associated with the gene and hence the gene is associated with the phenotype. For example, genes are selected that have variants associated with the phenotype within a 5 kb, 10 kb, 20 kb, 50 kb, or 100 kbwindow around the gene or have variants associated with the phenotype within 5 kb, 10 kb, 20 kb, 50 kb, or 100 kb of the transcribed region of the gene.

[0074] In example embodiments, variants are present in DNA regulatory elements (e.g., enhancers) active in a tissue associated with the phenotype of interest. In example embodiments, the regulatory elements regulate a gene, whereby the gene is associated with the phenotype. Any methods of linking enhancers comprising a variant to genes expressed in tissues can be used to identify genes associated with the phenotype of interest. In example embodiments, an Activity-by- Contact (ABC) model is used to link variants to genes. This model is based on the simple biochemical notion that an element’s quantitative effect on a gene should depend on its strength as an enhancer (“Activity”) weighted by how often it comes into 3D contact with the promoter of the gene (“Contact”), and that the relative contribution of an element on a gene’s expression should depend on the element’s effect divided by the total effect of all elements (see, e g., Fulco, et al. Activity-by-contact model of enhancer-promoter regulation from thousands of CRISPR perturbations. Nat Genet. 2019;51(12): 1664-1669. doi: 10.1038 / s41588-019-0538-0; and Moonen, et al., 2020, KLF4 Recruits SWI / SNF to Increase Chromatin Accessibility and Reprogram the Endothelial Enhancer Landscape under Laminar Shear Stress. bioRxiv 2020.07.10.195768, doi.org / 10. 1101 / 2020.07.10.195768). In example embodiments, an epigenome model, such as Roadmap, is used to link variants to genes to identify genes associated with a phenotype (see, e.g., Ernst, J., Kheradpour, P., Mikkelsen, T. et al. Mapping and analysis of chromatin state dynamics in nine human cell types. Nature 473, 43-49 (2011); Kundaje, A., Meuleman, W., Ernst, J. et al. Integrative analysis of 111 reference human epigenomes. Nature 518, 317-330 (2015); andegg2.wustl.edu / roadmap / web_portal / index.html). In example embodiments, an Enhancer-to-gene (E2G) strategy is a combined union of Activity-By-Contact and Roadmap Enhancer-to-gene (E2G) strategy (Roadmap-U-ABC E2G strategy) (see, e.g., US patent application publication US20210071255A1). In example embodiments, genes associated with phenotypes are selected using any method described in US Patent application publication US20210071255A1.Features For Use as Genomic Predictors

[0075] In example embodiments, once genes associated with a phenotype are selected correlation of genomic predictors for the genes is performed between two species and the amount of correlation can be used to predict concordance between the two species for the phenotype.Predicting concordance can indicate that any finding or modulating agent for the phenotype in the second species can be applied to the first species. In example embodiments, one or more genomic predictor is used to compare the selected genes. In example embodiments, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40, 50, 60, 70, 80, 90, or greater than 100 genomic predictors are used to predict concordance.

[0076] In example embodiments, a pair-wise ortholog score is used as a genomic predictor. In example embodiments, the conservation level of each gene between the species is determined. In example embodiments, the score is higher based on the most conversation between the genes in the two species (e.g., the genes are 100% conserved). Conserved sequences are identical or similar sequences in nucleic acids (DNA and RNA) or proteins (amino acids) across species (orthologous sequences) or within a genome (paralogous sequences). Different nucleic acid or amino acid sequences may be considered in the ortholog score. For example, in coding sequences, the nucleic acid and amino acid sequence may be conserved to different extents, as the degeneracy of the genetic code means that synonymous mutations in a coding sequence do not affect the amino acid sequence of its protein product. For example, amino acid sequences can be conserved to maintain the structure or function of a protein or domain. Conserved proteins undergo fewer amino acid replacements or are more likely to substitute amino acids with similar biochemical properties. Within a sequence, amino acids that are important for folding, structural stability, or that form a binding site may be more highly conserved.

[0077] In example embodiments, syntenic block harboring of each gene between the species is used as a genomic predictor. Analysis of synteny conservation between different genomes allows to identify similarity patterns and differences in genome structure and content. For example, for closely related genomes, synteny conservation can be performed at the DNA level, which can be useful to annotate newly sequenced genomes and to identify conserved non-coding sequences. For very distantly related genomes, detection of synteny conservation requires the development of statistical models or the construction of synteny profiles obtained from different genomes. For genomes sharing intermediate phylogenetic proximity, protein-coding genes may have retained enough sequence similarity and physical collinearity along chromosomes to allow synteny block reconstruction. Any method of calculating syntenic block harboring can be used (see, e.g., Drillon G, Carbone A, Fischer G. SynChro: a fast and easy tool to reconstruct and visualize synteny blocks along eukaryotic chromosomes. PLoS One. 2014;9(3):e92621; Wang Y, Tang H, Debarry ID, etal. MCScanX: a toolkit for detection and evolutionary analysis of gene synteny and collinearity. Nucleic Acids Res. 2012;40(7):e49; and Wang Y, Tang H, Wang X, Sun Y, Joseph PV, Paterson AH. Detection of colinear blocks and synteny and evolutionary analyses based on utilization of MCScanX. Nat Protoc. 2024;19(7):2206-2229).

[0078] In example embodiments, the number of paralogs for each gene in the two species is used as a genomic predictor. For example, a first species and a second species may have the same number of paralogs. In some embodiments, this suggests that the function of the genes is conserved between species. In another example the number of paralogs is different between two species, suggesting that two functions in one species may be performed by one or more paralogs or a function is lost in the species having less paralogs. In another example embodiment, the conservation between paralogs is a genomic predictor.

[0079] In example embodiments, the number regulatory elements associated with each gene between the species is used as a genomic predictor. In example embodiments, the regulatory elements are known to regulate the genes or are predicted to regulate the gene. In example embodiments, conservation of regulatory elements is a genomic predictor. In example embodiments, conservation of regulatory element sequence and position is a genomic predictor. In example embodiments, the chromatin modifications at the regulatory element are used as a genomic predictor (e.g., chromatin immunoprecipitation (ChIP) data). In example embodiments, the binding of transcription factors at the regulatory element is used as a genomic predictor. In example embodiments, the 3D chromatin architecture at the regulatory element is used as a genomic predictor (e.g., Hi-C). In example embodiments, the tissue or cell type specific activity of the regulatory element is used as a genomic predictor. In example embodiments, the chromatin accessibility of the regulatory element is used as a genomic predictor (e.g., ATAC-seq data).

[0080] In example embodiments, the number of tissues each selected gene is expressed in for each species is used as a genomic predictor. In example embodiments, the number of cell types each selected gene is expressed in for each species is used as a genomic predictor. In example embodiments, tissue expression for the selected genes is determined by use of a data base or determined experimentally. Example tissue and cell specific gene expression data bases are known, such as the Genotype-Tissue Expression (GTEx) Portal. The GTEx Portal is a comprehensive public resource for researchers studying tissue and cell-specific gene expression and regulation across individuals, development, and species (see, e.g., GTEx Analysis Release V8;GTEx Consortium. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. 2020;369(6509):1318-1330; and Kim-Hellmuth S, Aguet F, Oliva M, et al. Cell typespecific genetic regulation of gene expression across human tissues. Science. 2020;369(6509):eaaz8528).

[0081] In example embodiments, the connectivity, centrality, or position (i.e., protein interactions) in a protein-protein interaction (PPI) network for each gene in each species is used as a genomic predictor. In example embodiments, the conservation of connectivity, centrality, or position is determined for the genes in each species. In example embodiments, the number of proteins interacting with the gene in each species is determined.

[0082] In example embodiments, the number of functional processes identified for each gene is used as a genomic predictor. In example embodiments, gene set based enrichment analysis (GSEA) is used to determine functional processes for each gene. In example embodiments, the functional processes are ranked for the genes in each species and the top processes are compared to determine concordance. In example embodiments, only the top ranked functional process is compared between species. In example embodiments, any of the top 10 functional processes are compared. In example embodiments, the ranked list of all functional processes is analyzed by a machine learning algorithm as described further herein.

[0083] In example, embodiments, epigenetic profiles are compared for each gene between species. Epigenomic profiles have been described and are obtainable in databases (see, e.g., NIH Roadmap Epigenomics Mapping Consortium, ENCODE, Cistrome, and ChIP Atlas; ENCODE Project, Moore JE, Purcaro MJ, et al. Expanded encyclopaedias of DNA elements in the human and mouse genomes. Nature. 2020;583(7818):699-710; Li S, Wan C, Zheng R, et al. Cistrome- GO: a web server for functional enrichment analysis of transcription factor ChlP-seq peaks. Nucleic Acids Res. 2019;47(Wl):W206-W211; and Shinya Oki, Tazro Ohta, et al. ChIP- Atlas: a data-mining suite powered by full integration of public ChlP-seq data. EMBO Rep. (2018) e46255). The epigenomic profile can be a chromatin accessibility profile (e.g., ATAC-seq), a chromatin modification profile (e.g., ChlP-seq), a chromatin binding profile (e.g., ChlP-seq), a DNA methylation profile (e.g., Bisulfite-Seq), a DNase hypersensitivity profile (e.g., DNase-seq), or a DNA-DNA contact profile (e.g., Hi-C).

[0084] In example embodiments, data from a single cell atlas is used to determine gene expression, protein binding, chromatin modifications, chromatin accessibility for each gene inrelevant cell types for each species. As used herein “single cell atlas” refers to any collection of single cell data from any tissue sample of interest having a phenotype of interest (see, e.g., Rozenblatt-Rosen O, Stubbington MJT, Regev A, Teichmann SA., The Human Cell Atlas: from vision to reality., Nature. 2017 Oct 18;550(7677):451-453; and Regev, A. et al. The Human Cell Atlas Preprint available at bioRxiv at dx.doi.org / 10.1101 / 121202 (2017)).

[0085] In example embodiments, chromatin accessibility of the genomic region comprising each gene is analyzed using data obtained from Assay for Transposase Accessible Chromatin sequencing (ATAC-seq) (see, e.g., Buenrostro, J. D., Giresi, P. G., Zaba, L. C., Chang, H. Y., Greenleaf, W. J., Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nature methods 2013; 10 (12): 1213-1218; J. D. Buenrostro et al., Single-cell chromatin accessibility reveals principles of regulatory variation. Nature 523, 486-490 (2015); Cusanovich, D. A., Daza, R., Adey, A., Pliner, H., Christiansen, L., Gunderson, K. L., Steemers, F. J., Trapnell, C. & Shendure, J. Multiplex single-cell profiling of chromatin accessibility by combinatorial cellular indexing. Science. 2015 May 22;348(6237):910-4; and US20160208323A1; US20160060691A1; WO2017156336A1).

[0086] In example embodiments, the 3D chromatin architecture of the genomic region comprising each gene is analyzed using Hi-C (see, e.g., US patent Application publication US20240150830A1; U.S. patent application Ser. Nos. 15 / 532,353, 15 / 753,318, 16 / 308,386, 16 / 247,502, and 16 / 753,718; and International Patent Applications PCT / US2015 / 063272, PCT / US2016 / 047644, PCT / US2017 / 036649, PCT / US2018 / 054476, PCT / US2020 / 033436, PCT / US2020 / 064704; and Rao SS, Huntley MH, Durand NC, et al. A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell. 2014; 159(7): 1665- 1680). In example embodiments, chromatin looping at each gene is compared for cell types or tissues in each species. In example embodiments, protein binding sites at each gene is compared for cell types or tissues in each species.Predicting Concordance

[0087] In example embodiments, concordance or discordance is predicted for the phenotype between the first and second species based on the genomic predictors using a statistical analysis for individual genes. Statistical analysis can include any method described herein (e.g., nonparametric test (Wilcoxon rank sum test with continuity correction)). In example embodiments, machine learning algorithms are used to determine the concordance prediction (e.g., logisticregression (between the groups), multinomial logistic regression (within the group), and principal component analysis). For example, the machine learning algorithm can weigh the combination of genomic predictors to predict concordance. In example embodiments, the machine learning algorithm is trained using genomic predictor values between two species for genes associated with a phenotype that are experimentally shown to be concordant in the species. In example embodiments, the machine learning algorithm is trained using genomic predictor values between two species for genes associated with a phenotype that are experimentally shown to be discordant in the species. In example embodiments, the machine learning algorithm is trained on both concordant and discordant data. In example embodiments, linear regression is used to predict concordance based on the genomic predictors. In example embodiments, any dimensionality reduction method is used to predict concordance based on the genomic predictors (e.g., PCA).4. Examples

[0088] There are a variety implementations related to predicting concordance between two species and using a species as a reliable non-human animal model. Notwithstanding, with reference to FIGS. 1-17 and for the exemplary purposes of this disclosure, many of the implementations relate to predicting concordance and discordance in gene-phenotype relationships between a first species and a second species.

[0089] The following examples are to further illustrate the nature of the invention and are intended to be exemplary and non-limiting. It is believed that one of ordinary skill in the art can, using the preceding description and the following illustrative examples, make and utilize the present invention and practice the claimed methods. The following examples do not limit the invention and the scope of the invention is to be determined by the appended claims.

[0090] FIG. 1 shows genomic features referring to the distinctive characteristics or components of a genome (functional elements, regulatory elements, functional connections, epigenetic signature, etc.). Understanding the various genomic features and how they interact with each other is critical for understanding the function of genes and the development of diseases. In the analysis, Applicants removed predictor variables which were those meeting a cut-off of more than 40% of data missing, taking around 25% of the original data (0,0: 6113, 0,1 : 148, 1,0: 27 and 1,1 : 69) into account for further analysis (including 12 predictors).

[0091] FIGS. 2A-2B presents a comparison of regression models around the predictor variable. FIG. 2A compares the (1,1) and (0,0) groups and FIG. 2B compares the (0,1) and (1,0) groups. Regression analysis was used to investigate the relationship among predictor variables and look for variations between two groups. Equation of a simple logistic regression (LR) to contrast genes between groups:One predictor Model: glm(formula = Group ~ Predictor 1-18, family = “binomial”, data = data).Multi Predictor Model: glm(formula = Group ~ Predictor 1, Predictor 2 , Predictor 12, family = “binomial”, data = data).The models consistently indicate a significant association of two predictor variables (Human Tissue count and Human GO term count) in the dataset.

[0092] FIG. 3 shows the effect size refers to the magnitude of the relationship between the predictor variables and the response variables.

[0093] FIG. 4 presents a description of the Data-set. GenAge is an ageing gene database which includes a curated database of over 300 genes related to human ageing and a database of over 2000 ageing- and longevity-associated genes in model organisms (yeast, worms, flies, mice, etc.). 307 human genes are known to be linked to human longevity. 136 genes are known to influence longevity in a mouse. The figure shows overlap between the human and mouse genes.

[0094] FIG. 5 and FIGS. 6A-6E illustrate that pair-wise correlation is important to understand the relationship between each pair of variables in the dataset. FIG. 5 shows the proportion missing for certain genes. FIGS. 6A-6E presents pair-wise correlation (Pearson Correlation coefficients with a P value cut-off of < 0.05) of all 18 genomic predictors based on genes detected in each of the four groups is shown.

[0095] FIGS. 7A-7C illustrate principal component analysis (PCA) among the predictor variables. PCA was applied to identifying linear combinations of the predictor variables that capture the maximum amount of variation in the dataset. A PCA plot illustrating the variance in the dataset using 12 different predictor variables across all four groups is presented. Cos2represents the proportion of the variance of a given variable that is explained by a particular principal component.

[0096] FIG. 8 shows results from use of the MDMR method to analyze the relationship between multiple predictor variables and groups of observations in the dataset. The MDMR method was used to analyze different random set of genes from 00 category to check the consistency in the analysis shown in FIG. 9.

[0097] FIG. 10 shows that gene set based enrichment analysis (GSEA) contrasts genes with, c. ., (1,1) vs. (0,0) designations. The gene set used to generate the multi-predictor LR model was analyzed using the Gene Ontology (GO-BP) and pathways (KEGG, Reactome and WikiPathway) databases. The gene sets present in the categories are presented below.GO-BP enriched by gene-set present in El category• G0:0070887 - cellular response to chemical stimulus• GO: 1901700 - response to oxygen-containing compound• GO:0031325 - positive regulation of cellular metabolic process• G0:0042221 - response to chemical• GO: 0009893 - positive regulation of metabolic process• GO: 0010604 - positive regulation of macromolecule metabolic process• G0:0051716 - cellular response to stimulus• GO: 0033554 - cellular response to stress• GO: 0048583 - regulation of response to stimulus• G0:0048584 - positive regulation of response to stimulusGO-BP enriched by gene-set present in 0.0 category• G0:0050794 - regulation of cellular process• G0:0032502 - developmental process• G0:0048856 - anatomical structure development• G0:0065007 - biological regulation• G0:0050789 - regulation of biological process• G0:0007275 - multicellular organism development• G0:0032501 - multicellular organismal process• G0:0048731 - system development• G0:0048518 - positive regulation of biological process• G0:0009653 - anatomical structure morphogenesis

[0098] A description of some data sets used for cross-species orthologs analysis are presented below.• Genes correlated (DE) with non-human species MLS (J. Yuyang Lu et al., 2022): Correlation statistics (Spearman and BH adjusted P values) of 14,773 non-human (mouse) genes associated with MLS (maximum life span) in 26 different non-human species before and after the body mass (BM) correction.• Gene Orthologs Mapping: 1 : 1 orthologous gene mappings based on the information retrieved from the Ensembl database (WORMHOLE - www.jax.org / research-and- faculty / resources / wormhole).• Genes significantly correlated with MLS before BM correction: 8011 genes (applying BH adjusted P value cut-off < 0.05).• Genes significantly correlated with MLS after BM correction: 7204 genes (applying BH adjusted P value cut-off < 0.05).• Genes associated (GWAS) with human longevity (Timmers PR et a / ., 2019): GWAS summary statistics for 7,12,956 SNPs (applying P value cut-off < 0.05).• SNP to Gene Mapping: Gene-based association test using GWAS Summary Statistics (genotype reference data from UK Biobank). 17,392 genes were identified by including a 10 kb window before / after a gene for the SNP mapping (cran.r- project.org / web / packages / snpsettest / index.html).• Genes associated (GWAS) with human longevity (Deelen J et al., 2019): GWAS summary statistics for 1,52,790 SNPs (applying P value cut-off < 0.05).• SNP to Gene Mapping: Gene-based association test using GWAS Summary Statistics (genotype reference data from 1000 Genomes phase 3). 8,695 genes were identified by including a 10 kb window before / after a gene for the SNP mapping (cran.r- project.org / web / packages / snpsettest / index.html).

[0099] FIG. 11 shows cross-species aging genes comparison for human and non-human species. Cross-species ortholog comparison of the genes in human and non-human species:1. Conservation level (pair-wise ortholog score).2. Syntenic block harboring the gene between the species.3. Number of Paralogs.4. Number of known regulatory elements (TFs).5. Expression patterns of the gene (number of tissues).6. Position in PPI network (connectivity / centrality metrics).7. Number of functional processes (Gene Ontologies).Types of analysis were:1. Non-parametric test (Wilcoxon rank sum test with continuity correction).2. Logistic regression (between the groups).3. Multinomial logistic regression (within the group).4. Principal Component Analysis.

[0100] FIGS. 12A-12C show cross-species aging gene comparison for the strength of the association in humans and other non-human species. The results show highly significant genes from GWAS of human longevity (highlighted in the figures) show weaker correlation with non- human species MLS.

[0101] FIGS. 13A-13D show cross-species ageing genes comparison of a gene-set comparison. Gene sets were as follows:• “11” or “1,1”=Significant in both (non-human and human)• “10” or “1,0” = Significant in non-human but non-significant in human• “01” or “0,1” = Non-significant in non-human but significant in human• “00” or “1,1” = Non-significant in both (non-human and human) BM correlations are shown for each group.

[0102] FIGS. 14A-14H presents pair-wise correlation of the predictors within each group with Pearson Correlation coefficients with P value cut-off < 0.05. P-Value is the probability that one would have found the current result if the correlation coefficient were in fact zero (null hypothesis). If the probability is lower than 5% (P<0.05), the correlation coefficient is considered statistically significant.

[0103] FIGS. 15A-15B presents regression model results to contrast genes between groups. The analysis removed the predictors with a cut-off of over 40% of data missing and excluded NA's explicitly before the fitting of a model. Below equation of a simple logistic regression (LR) to contrast genes with, e.g., (1,1) vs. (0,0) designations: core) + 02(Human_paralogs_count) +03(Non-human_paralogs_count) + P4(Human_PPl_Connections) + 0s(Non- human PPI Connections) + 06(Human_tissue_Count) + 07(Non-human_tissue_Count) +08(Human_GO_terms) +09(Non-human_GO_terms) +0io(Human_Syntenic_block) + 0n(Non- human_ Syntenic block) + s

[0104] Table 1 presents a summary table showing concordance / discordance of the predictor variables.TABLE 1

[0105] FIGS. 16A-16D show cross-species aging genes comparison for a principal component analysis. A high cos2 indicates a good representation of a predictor on the principal component. In FIGS. 16A-16D, the “0,0” designation indicates non-significant gene sets in both non-humans and humans; the “1,1” designation indicates significant gene sets in both (non-humans and humans); the “1,0” designation indicates significant gene sets in non-humans but non-significantgene sets in humans; and “0,1” designation indicates non-significant gene sets in non-humans but significant gene sets in humans.

[0106] FIGS. 17A-17D show human tissue specific gene expression comparison within the groups — distribution of Tau index. The tau index indicates how specific or broadly a gene expressed within studied tissues (tissue-specific gene expression data from GTEx Version 8). Genes with a tau index close to 1 are more specifically expressed in one tissue, while genes with a tau index closer to 0 are equally expressed across all tissues studied. In comparison to non-human species, human longevity associated genes are more tissue-specific.

[0107] The results presented above allow the methods to easily be extended to other phenotypes and explore additional genomic features.

[0108] As required, detailed embodiments of the present invention are disclosed herein. It is to be understood that the disclosed embodiments are merely exemplary of the invention that may be embodied in various forms. Therefore, specific structural and functional details disclosed herein are not to be interpreted as limits, but merely as a basis for teaching one skilled in the art to employ the present invention. The specific examples below will enable the invention to be better understood. However, they are given merely by way of guidance and do not imply any limitation.

[0109] It is to be appreciated that certain features of the invention, which are, for clarity, described herein in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the invention that are, for brevity, described in the context of a single embodiment, may also be provided separately or in any subcombination. Further reference to values stated in ranges includes each and every value and combination of values within that range.5. Paragraphs

[0110] The invention is illustrated by the following numbered paragraphs:1. A method of predicting concordance and discordance in gene-phenotype relationships between a first species and a second species, the method comprising: a) selecting one or more genes associated with a phenotype, wherein the selected genes are associated with the phenotype (a) in both the first species and the second species, (b) in the first species only, and / or (c) in the second species only;b) determining a pair-wise correlation of a plurality of genomic predictors for the one or more selected genes in the first species compared to the second species; and c) predicting the concordance or discordance for the phenotype between the first and second species based on the genomic predictors using a statistical analysis for individual genes.2. The method of paragraph 1, wherein the plurality of genomic predictors comprise: a pair-wise ortholog score, whereby the conservation level of each gene between the species is determined, syntenic block harboring of each gene between the species, the number of paralogs of each gene in each species, the number of regulatory elements associated with each gene in each species, number of tissues each gene is expressed in for each species, whereby expression patterns of each gene is determined, the connectivity, centrality, or position in a protein-protein interaction (PPI) network for each gene in each species, and / or the number of functional processes identified for each gene.3. The method of paragraph 1 or 2, wherein the first species is human and the second species is a non-human animal.4. The method of any one of paragraphs 1-3, wherein the statistical analysis is a pair-wise t- test.5. The method of any one of paragraphs 1-4, further comprising calculating the probability of the concordance or discordance based on the genomic predictors for each gene using a multipredictor linear regression model.6. The method of any one of paragraphs 1-5, wherein the phenotype is longevity.7. The method of any one of paragraphs 1-5, wherein the phenotype is a disease phenotype.8. The method of any one of paragraphs 1-7, wherein the one or more genes are selected from the Gen Age database.9. The method of any one of paragraphs 1-7, wherein the one or more genes are selected from genes linked to the phenotype based on genome-wide association studies (GWAS) for the phenotype.10. The method of any one of paragraphs 1-9, wherein the one or more selected genes are all expressed in the same tissue or cell type.11 . The method of any one of paragraphs 1-10, further comprising removing from the analysis any genomic predictors for which a threshold amount of data is missing from the database.12. The method of anyone of paragraphs 1-12, additionally comprising removing from analysis the genomic predictors having a threshold amount of data missing for the one or more genes to ensure that there are enough genomic predictors for each gene to be confident in the predictions about concordant or discordant effects of the phenotype between the species.13. The method of paragraph 12, wherein the threshold amount is more than 40% of data being missing.14. The method of any one of paragraphs 1-13, wherein the pair-wise correlation comprises a Pearson Correlation coefficient to determine correlation between the genomic predictors between the first species and the second species.15. The method of any one of paragraphs 1-14, wherein dimension reduction is used to analyze the pair-wise correlations to identify combinations of the genomic predictors that have the most predictive power.16. The method of paragraph 15, wherein the dimension reduction comprises a principal component analysis (PCA), whereby linear combinations of the genomic predictors that have the most predictive power are identified.17. The method of any one of paragraphs 1-16, wherein a linear regression model is used in the concordance or discordance analysis.18. The method of paragraph 17, wherein predicted probability is used to evaluate the accuracy of the linear regression model.19. The method of any one of paragraphs 1-18, wherein a multivariate distance matrix regression (MDMR) method is used to analyze the relationship between multiple genomic predictors and the one or more genes.20. The method of paragraph 19, further comprising using the MDMR method to analyze different random sets of genes from a group of genes present in neither the first species nor second species to check for consistency of analysis.21. The method of any one of paragraphs 1-20, wherein gene set based enrichment analysis (GSEA) is used to identify functional processes in the concordance or discordance analysis.22. The method of any one of paragraphs 1-21, wherein a tau index is used in the concordance or discordance analysis, wherein the tau index indicates how specific or broadly each gene isexpressed across tissues.23. The method of any one of paragraphs 3-22, further comprising selecting the second species as a non-human animal model to test a medicament for modulating the phenotype in the first species if concordance for the phenotype between the first and second species is predicted.24. The method of any one of paragraphs 1-23, wherein the one or more genes associated with a phenotype are selected by determining genes associated with the phenotype in the first and second species; and selecting genes (a) present in both the first species and the second species or (b) present in the first species only.25. A method of identifying an agent or target gene for modulating a phenotype in a subject comprising contacting a non-human animal model with one or more agents specific for targeting one or more genes, wherein the phenotype modulated in the non-human animal model is concordant with the phenotype in a human subject, wherein concordance of the phenotype is determined by a method comprising: a) selecting one or more genes associated with a phenotype, wherein the selected genes are associated with the phenotype (a) in both the first species and the second species, (b) in the first species only, and / or (c) in the second species only; b) determining a pair-wise correlation of a plurality of genomic predictors for the one or more selected genes in the first species compared to the second species; and c) predicting the concordance or discordance for the phenotype between the first and second species based on the genomic predictors using a statistical analysis for individual genes.26. The method of paragraph 25, wherein the plurality of genomic predictors comprise: a pair-wise ortholog score, whereby the conservation level of each gene between the species is determined, syntenic block harboring of each gene between the species, the number of paralogs of each gene in each species, the number of regulatory elements associated with each gene in each species, number of tissues each gene is expressed in for each species, whereby expression patterns of each gene is determined, the connectivity, centrality, or position in a protein-protein interaction (PPI) network for each gene in each species, and / or the number of functional processes identified for each gene.27. The method of paragraph 25 or 26, wherein the first species is human and the second species is a non-human animal.28. The method of any one of paragraphs 25-27, wherein the statistical analysis is a pair-wise t-test.29. The method of any one of paragraphs 25-28, further comprising calculating the probability of the concordance or discordance based on the genomic predictors for each gene using a multipredictor linear regression model.30. The method of any one of paragraphs 25-29, wherein the phenotype is longevity.31. The method of any one of paragraphs 25-29, wherein the phenotype is a disease phenotype.32. The method of any one of paragraphs 25-31, wherein the one or more genes are selected from the Gen Age database.33. The method of any one of paragraphs 25-31, wherein the one or more genes are selected from genes linked to the phenotype based on genome-wide association studies (GWAS) for the phenotype.34. The method of any one of paragraphs 25-33, wherein the one or more selected genes are all expressed in the same tissue or cell type.35. The method of any one of paragraphs 25-34, further comprising removing from the analysis any genomic predictors for which a threshold amount of data is missing from the database.36. The method of anyone of paragraphs 25-35, additionally comprising removing from analysis the genomic predictors having a threshold amount of data missing for the one or more genes to ensure that there are enough genomic predictors for each gene to be confident in the predictions about concordant or discordant effects of the phenotype between the species.37. The method of paragraph 36, wherein the threshold amount is more than 40% of data being missing.38. The method of any one of paragraphs 25-37, wherein the pair-wise correlation comprises a Pearson Correlation coefficient to determine correlation between the genomic predictors between the first species and the second species.39. The method of any one of paragraphs 25-38, wherein dimension reduction is used to analyze the pair-wise correlations to identify combinations of the genomic predictors that have the most predictive power.40. The method of paragraph 39, wherein the dimension reduction comprises a principalcomponent analysis (PCA), whereby linear combinations of the genomic predictors that have the most predictive power are identified.41. The method of any one of paragraphs 25-40, wherein a linear regression model is used in the concordance or discordance analysis.42. The method of paragraph 41 , wherein predicted probability is used to evaluate the accuracy of the linear regression model.43. The method of any one of paragraphs 25-42, wherein a multivariate distance matrix regression (MDMR) method is used to analyze the relationship between multiple genomic predictors and the one or more genes.44. The method of paragraph 43, further comprising using the MDMR method to analyze different random sets of genes from a group of genes present in neither the first species nor second species to check for consistency of analysis.45. The method of any one of paragraphs 25-44, wherein gene set based enrichment analysis (GSEA) is used to identify functional processes in the concordance or discordance analysis.46. The method of any one of paragraphs 25-45, wherein a tau index is used in the concordance or discordance analysis, wherein the tau index indicates how specific or broadly each gene is expressed across tissues.47. The method of any one of paragraphs 25-46, wherein the one or more genes associated with a phenotype are selected by determining genes associated with the phenotype in the first and second species; and selecting genes (a) present in both the first species and the second species or (b) present in the first species only.6. Aspects[0U1] The invention is illustrated by the following aspects:Aspect 1. A method of predicting concordance and discordant gene-phenotype relationships between a first species and a second species, the method comprising: characterizing genes into four groups based on the presence or absence of a gene-phenotype relationship in the first species and the second species, the four groups being (a) present in both the first species and the second species, (b) present in the first species only, (c) present in the second species only, and (d) present in neither the first species nor second species;curating a plurality of genomic predictors of the concordance / discordance of the genephenotype relationships associated with each of the four groups; conducting a pair-wise correlation plot of the plurality of genomic predictors; and predicting the genetic concordance or discordance reflected in the four groups based on the genomic predictors using a statistical analysis for individual genes.Aspect 2. The method of Aspect 1, wherein the first species is human.Aspect 3. The method of Aspect 1 or Aspect 2, wherein the statistical analysis is a pair-wise t-test.Aspect 4. The method of any one of Aspects 1-3, further comprising calculating the probability of the concordance or discordance reflected in the four groups based on the genomic predictors for each gene using a multi-predictor linear regression model between two or more of the four groups.Aspect 5. The method of any one of Aspects 1-4, wherein the characteristic is longevity.Aspect 6. The method of any one of Aspects 1-5, wherein the genes are selected from theGenAge database.Aspect 7. The method of any one of Aspects 6, further comprising removing from analysis any genomic predictors for which a threshold amount of data is missing from the database.Aspect 8. The method of anyone of Aspect 7, additionally comprising removing from analysis the predictor variables having a threshold amount of data is missing from the database to ensure that there are enough genomic predictors for each gene to be confident in the predictions about concordant or discordant effects of the gene across the species studied.Aspect 9. The method of Aspect 7, wherein the threshold amount is more than 40% of data being missing.Aspect 10. The method of any one of Aspects 1-9, wherein the pair-wise correlation comprises a Pearson Correlation coefficient to determine correlation between the genomic predictors between the first species and the second species.Aspect 11. The method of any one of Aspects 1-9, wherein the pair- wise correlation plot comprises a principal component analysis (PCA) is used to identify linear combinations of predictor variables.Aspect 12. The method of any one of Aspects 1-9, wherein a regression model is used in the concordance or discordance analysis.Aspect 13. The method of Aspect 12, wherein predicted probability was used to evaluate the accuracy of the linear regression model.Aspect 14. The method of any one of Aspects 1-9, wherein a multivariate distance matrix regression (MDMR) was used to analyze the relationship between multiple predictor variables and groups.Aspect 15. The method of Aspect 14, further comprising using the MDMR method to analyze different random sets of genes from a group of genes present in neither the first species nor second species to check for consistency of analysis.Aspect 16. The method of any one of Aspects 1-9, wherein gene set based enrichment analysis (GSEA) was used in the concordance or discordance analysis.Aspect 17. The method of any one of Aspects 1-9, further comprising a human tissue specific gene expression comparison within at least one of the four groups using a Tau index.Aspect 18. The method of any one of Aspects 1-17, wherein the first species is a human, and the second species is a non-human animal, and wherein the characteristic is longevity.Aspect 19. A method of selecting a first species for pharmaceutical testing of a medicament for use with a second species, the method comprising: characterizing genes of a plurality of species into four groups based on the presence or absence of a gene-phenotype relationship in the first species and the second species, the four groups being (a) present in both the first species and the second species, (b) present in the first species only, (c) present in the second species only, and (d) present in neither the first species nor second species; curating a plurality of genomic predictors of the concordance / discordance of the genephenotype relationships associated with each of the four groups; conducting a pair-wise correlation plot of the plurality of genomic predictors; predicting the concordance or discordance reflected in the four groups based on the genomic predictors using a statistical analysis for individual genes; and selecting the first species based on concordance of a preselected gene between the first and second species.Aspect 20. The method of Aspect 19, wherein the statistical analysis is a pair-wise t-test for individual genes in each species.Aspect 21. The method of Aspect 19 or Aspect 20, wherein the second species is human.Aspect 22. The method of any one of Aspects 19-21, further comprising calculating the probability of the concordance or discordance reflected in the four groups based on the genomic predictors for each gene using a multi-predictor linear regression model between two or more of the four groups.Aspect 23. The method of any one of Aspects 19-22, wherein the genes are selected from the GenAge database.Aspect 24. The method of Aspect 23, further comprising removing from analysis any genomic predictors for which a threshold amount of data is missing from the database.Aspect 25. The method of Aspect 24, additionally comprising removing from analysis the predictor variables having a threshold amount of data is missing from the database to ensure that there are enough genomic predictors for each gene to be confident in the predictions about concordant or discordant effects of the gene across the species studied.Aspect 26. The method of Aspect 25, wherein the threshold amount is more than 40% of data being missing.Aspect 27. The method of any one of Aspects 19-26, wherein the pair-wise correlation comprises a Pearson Correlation coefficient to determine correlation between the genomic predictors between the first species and the second species.Aspect 28. The method of any one of Aspects 19-26, wherein the pair-wise correlation plot comprises a principal component analysis (PCA) is used to identify linear combinations of predictor variables.Aspect 29. The method of any one of Aspects 19-26, wherein a regression model is used to predict concordance or discordance analysis.Aspect 30. The method of Aspect 29, wherein predicted probability was used to evaluate the accuracy of the linear regression model.Aspect 31. The method of any one of Aspects 19-26, wherein a multivariate distance matrix regression (MDMR) was used to analyze the relationship between multiple predictor variables and groups.Aspect 32. The method of Aspect 31, further comprising using the MDMR method to analyze different random sets of genes from a group of genes present in neither the first species nor second species to check for consistency of analysis.Aspect 33. The method of any one of Aspects 19-26, wherein gene set based enrichment analysis (GSEA) was used in the concordance or discordance analysis.Aspect 34. The method of any one of Aspects 19-26, further comprising a human tissue specific gene expression comparison within at least one of the four groups using a Tau index.Aspect 35. The method of any one of Aspects 19-34, wherein the first species is a human, and the second species is a non-human animal, and wherein the characteristic is longevity.

[0112] Many additional implementations are possible. Further implementations are within the CLAIMS. In places where the description above refers to particular methods, it should be readily apparent that a number of modifications may be made without departing from the spirit thereof and that these methods may be applied to other implementations disclosed or undisclosed. The presently disclosed methods are, therefore, to be considered in all respects as illustrative and not restrictive.

Claims

CLAIMSWe Claim:

1. A method of predicting concordance and discordance in gene-phenotype relationships between a first species and a second species, the method comprising: a) selecting one or more genes associated with a phenotype, wherein the selected genes are associated with the phenotype (a) in both the first species and the second species, (b) in the first species only, and / or (c) in the second species only; b) determining a pair-wise correlation of a plurality of genomic predictors for the one or more selected genes in the first species compared to the second species; and c) predicting the concordance or discordance for the phenotype between the first and second species based on the genomic predictors using a statistical analysis for individual genes.

2. The method of claim 1, wherein the plurality of genomic predictors comprise: a pair-wise ortholog score, whereby the conservation level of each gene between the species is determined, syntenic block harboring of each gene between the species, the number of paralogs of each gene in each species, the number of regulatory elements associated with each gene in each species, number of tissues each gene is expressed in for each species, whereby expression patterns of each gene is determined, the connectivity, centrality, or position in a protein-protein interaction (PPI) network for each gene in each species, and / or the number of functional processes identified for each gene.

3. The method of claim 1 or 2, wherein the first species is human and the second species is a non-human animal.

4. The method of any one of claims 1-3, wherein the statistical analysis is a pair-wise t-test.

5. The method of any one of claims 1-4, further comprising calculating the probability of the concordance or discordance based on the genomic predictors for each gene using a multi -predictor linear regression model.

6. The method of any one of claims 1-5, wherein the phenotype is longevity.

7. The method of any one of claims 1-5, wherein the phenotype is a disease phenotype.

8. The method of any one of claims 1-7, wherein the one or more genes are selected from the GenAge database.

9. The method of any one of claims 1-7, wherein the one or more genes are selected from genes linked to the phenotype based on genome-wide association studies (GWAS) for the phenotype.

10. The method of any one of claims 1-9, wherein the one or more selected genes are all expressed in the same tissue or cell type.

11. The method of any one of claims 1-10, further comprising removing from the analysis any genomic predictors for which a threshold amount of data is missing from the database.

12. The method of any one of claims 1-11, additionally comprising removing from analysis the genomic predictors having a threshold amount of data missing for the one or more genes to ensure that there are enough genomic predictors for each gene to be confident in the predictions about concordant or discordant effects of the phenotype between the species.

13. The method of claim 12, wherein the threshold amount is more than 40% of data being missing.

14. The method of any one of claims 1-13, wherein the pair-wise correlation comprises a Pearson Correlation coefficient to determine correlation between the genomic predictors between the first species and the second species.

15. The method of any one of claims 1-14, wherein dimension reduction is used to analyze the pair- wise correlations to identify combinations of the genomic predictors that have the most predictive power.

16. The method of claim 15, wherein the dimension reduction comprises a principal component analysis (PCA), whereby linear combinations of the genomic predictors that have the most predictive power are identified.

17. The method of any one of claims 1-16, wherein a linear regression model is used in the concordance or discordance analysis.

18. The method of claim 17, wherein predicted probability is used to evaluate the accuracy of the linear regression model.

19. The method of any one of claims 1-18, wherein a multivariate distance matrix regression (MDMR) method is used to analyze the relationship between multiple genomic predictors and the one or more genes.

20. The method of claim 19, further comprising using the MDMR method to analyze different random sets of genes from a group of genes present in neither the first species nor second species to check for consistency of analysis.

21. The method of any one of claims 1 -20, wherein gene set based enrichment analysis (GSEA) is used to identify functional processes in the concordance or discordance analysis.

22. The method of any one of claims 1-21, wherein a tau index is used in the concordance or discordance analysis, wherein the tau index indicates how specific or broadly each gene is expressed across tissues.

23. The method of any one of claims 3-22, further comprising selecting the second species as a non-human animal model to test a medicament for modulating the phenotype in the first species if concordance for the phenotype between the first and second species is predicted.

24. The method of any one of claims 1-23, wherein the one or more genes associated with a phenotype are selected by determining genes associated with the phenotype in the first and second species; and selecting genes (a) present in both the first species and the second species or (b) present in the first species only.

25. A method of identifying an agent or target gene for modulating a phenotype in a subject comprising contacting a non-human animal model with one or more agents specific for targeting one or more genes, wherein the phenotype modulated in the non-human animal model is concordant with the phenotype in a human subject, wherein concordance of the phenotype is determined by a method comprising: a) selecting one or more genes associated with a phenotype, wherein the selected genes are associated with the phenotype (a) in both the first species and the second species, (b) in the first species only, and / or (c) in the second species only; b) determining a pair-wise correlation of a plurality of genomic predictors for the one or more selected genes in the first species compared to the second species; and c) predicting the concordance or discordance for the phenotype between the first and second species based on the genomic predictors using a statistical analysis for individual genes.

26. The method of claim 25, wherein the plurality of genomic predictors comprise: a pair-wise ortholog score, whereby the conservation level of each gene between the species is determined, syntenic block harboring of each gene between the species, the number of paralogs of each gene in each species, the number of regulatory elements associated with each gene in each species, number of tissues each gene is expressed in for each species, whereby expression patterns of each gene is determined, the connectivity, centrality, or position in a protein-protein interaction (PPI) network for each gene in each species, and / or the number of functional processes identified for each gene.

27. The method of claim 25 or 26, wherein the first species is human and the second species is a non-human animal.

28. The method of any one of claims 25-27, wherein the statistical analysis is a pair-wise t- test.

29. The method of any one of claims 25-28, further comprising calculating the probability of the concordance or discordance based on the genomic predictors for each gene using a multipredictor linear regression model.

30. The method of any one of claims 25-29, wherein the phenotype is longevity.

31. The method of any one of claims 25-29, wherein the phenotype is a disease phenotype.

32. The method of any one of claims 25-31, wherein the one or more genes are selected from the Gen Age database.

33. The method of any one of claims 25-31, wherein the one or more genes are selected from genes linked to the phenotype based on genome-wide association studies (GWAS) for the phenotype.

34. The method of any one of claims 25-33, wherein the one or more selected genes are all expressed in the same tissue or cell type.

35. The method of any one of claims 25-34, further comprising removing from the analysis any genomic predictors for which a threshold amount of data is missing from the database.

36. The method of anyone of claims 25-35, additionally comprising removing from analysis the genomic predictors having a threshold amount of data missing for the one or more genes to ensure that there are enough genomic predictors for each gene to be confident in the predictions about concordant or discordant effects of the phenotype between the species.

37. The method of claim 36, wherein the threshold amount is more than 40% of data being missing.

38. The method of any one of claims 25-37, wherein the pair-wise correlation comprises a Pearson Correlation coefficient to determine correlation between the genomic predictors between the first species and the second species.

39. The method of any one of claims 25-38, wherein dimension reduction is used to analyze the pair-wise correlations to identify combinations of the genomic predictors that have the most predictive power.

40. The method of claim 39, wherein the dimension reduction comprises a principal component analysis (PCA), whereby linear combinations of the genomic predictors that have the most predictive power are identified.

41. The method of any one of claims 25-40, wherein a linear regression model is used in the concordance or discordance analysis.

42. The method of claim 41, wherein predicted probability is used to evaluate the accuracy of the linear regression model.

43. The method of any one of claims 25-42, wherein a multivariate distance matrix regression (MDMR) method is used to analyze the relationship between multiple genomic predictors and the one or more genes.

44. The method of claim 43, further comprising using the MDMR method to analyze different random sets of genes from a group of genes present in neither the first species nor second species to check for consistency of analysis.

45. The method of any one of claims 25-44, wherein gene set based enrichment analysis (GSEA) is used to identify functional processes in the concordance or discordance analysis.

46. The method of any one of claims 25-45, wherein a tau index is used in the concordance or discordance analysis, wherein the tau index indicates how specific or broadly each gene is expressed across tissues.

47. The method of any one of claims 25-46, wherein the one or more genes associated with a phenotype are selected by determining genes associated with the phenotype in the first and second species; and selecting genes (a) present in both the first species and the second species or (b) present in the first species only.

Images