Methods and compositions for imputing or predicting genotype or phenotype

1. A method of parametrically representing genotypic or phenotypic data obtained from one or more plant populations to impute or predict a genotype or phenotype of interest associated with an agronomic trait, the method comprising:

providing genotypic or phenotypic data obtained from one or more plant populations;

generating a universal latent space representation by encoding discrete or continuous variables derived from genotypic or phenotypic data into latent vectors through a machine learning-based encoder framework in a computer system, wherein the latent space representation is independent of the underlying genotypic or phenotypic data;

decoding the latent space representation by a decoder, thereby imputing or predicting the genotype or phenotype of interest by the decoded latent representation; and

selecting one or more plant populations or members thereof based on the imputed or predicted genotype or phenotype of interest, wherein the imputed or predicted genotype or phenotype of interest is associated with an agronomic trait.

2. The method of claim 1 , wherein the genotypic data comprises a collection of genotypic markers or single nucleotide polymorphisms (SNPs).

3. The method of claim 2 , wherein a subset of the genotypic markers or SNPs are from a plurality of genetically divergent plant populations.

4. The method of claim 1 , wherein a subset of the discrete variables is a plurality of genotypic markers or SNPs localized to a chromosomal segment.

5. The method of claim 1 , wherein the encoder framework is based on a neural network algorithm.

6. The method of claim 1 , wherein the imputed or predicted genotype of interest is a plurality of haplotypes.

7. The method of claim 1 , wherein the decoder imputes or predicts local high-density (HD) SNPs.

8. The method of claim 1 , wherein the genotypic data is obtained from populations of plants derived from two or more breeding programs, wherein the breeding programs do not comprise an identical set of genotypic markers or SNPs corresponding to the genotypic data.

9. The method of claim 1 , wherein the decoder imputes or predicts local high-density (HD) SNPs of one population based on the decoding of genotypic data of another population.

10. The method of claim 1 , wherein the decoder imputes or predicts haplotypes for one population based on the decoding of genotypic data of another population.

11. The method of claim 1 , wherein the decoder imputes or predicts population coancestry for one or more populations.

12. The method of claim 1 , the method comprising: training the decoder to learn a prediction or imputation of a genotype or phenotype of interest based on an objective function for the encoded latent vectors.

13. The method of claim 12 , the method further comprising: decoding by the decoder the encoded latent vector for the objective function.

14. The method of claim 13 , the method comprising: providing an output for the objective function of the decoded latent vector.

15. The method of claim 1 , the method comprising crossing with another population or member the one or more selected populations or members thereof imputed or predicted to comprise a genotype of interest, wherein the genotype of interest is associated with a desirable agronomic trait.

16. The method of claim 1 , the method comprising counter-selecting from a breeding program the one or more selected populations or members thereof imputed or predicted to comprise a genotype of interest, wherein the genotype of interest is associated with an undesirable agronomic trait.

17. The method of claim 1 , the method comprising crossing with another population or member the one or more selected populations or members thereof imputed or predicted to comprise a phenotype of interest, wherein the phenotype of interest is associated with a desirable agronomic trait.

18. The method of claim 1 , the method comprising counter-selecting from a breeding program the one or more selected populations or members thereof imputed or predicted to comprise a phenotype of interest, wherein the phenotype of interest is associated with an undesirable agronomic trait.

19. The method of claim 1 , wherein the plant is a soybean, maize, sorghum, cotton, canola, sunflower, rice, wheat, sugarcane, alfalfa tobacco, barley, cassava, peanuts, millet, oil palm, potatoes, rye, or sugar beet plant.

20. The method of claim 1 , wherein the predicted phenotype of interest is yield gain.

21. The method of claim 1 , wherein the imputed or predicted phenotype of interest is root lodging, stalk lodging, brittle snap, ear height, grain moisture, plant height, disease resistance, drought tolerance, or a combination thereof.

22. The method of claim 1 , wherein the decoder imputes or predicts a molecular phenotype selected from gene expression, chromatin accessibility, DNA methylation, histone modifications, recombination hotspot, genomic landing locations for transgenes, transcription factor binding status, or a combination thereof.

23. A computing device comprising a processor configured to perform the steps of the method: providing genotypic or phenotypic data obtained from one or more plant populations; generating a universal latent space representation by encoding discrete or continuous variables derived from genotypic or phenotypic data into latent vectors through a machine learning-based encoder framework in a computer system, wherein the latent space representation is independent of the underlying genotypic or phenotypic data; decoding the latent space representation by a decoder, thereby imputing or predicting the genotype or phenotype of interest by the decoded latent representation; and selecting one or more plant populations or members thereof based on the imputed or predicted genotype or phenotype of interest, wherein the imputed or predicted genotype or phenotype of interest is associated with an agronomic trait.

24. A computer-readable medium comprising instructions which, when executed by a computing device, cause the computing device to carry out the steps of: providing genotypic or phenotypic data obtained from one or more plant populations; generating a universal latent space representation by encoding discrete or continuous variables derived from genotypic or phenotypic data into latent vectors through a machine learning-based encoder framework in a computer system, wherein the latent space representation is independent of the underlying genotypic or phenotypic data; decoding the latent space representation by a decoder, thereby imputing or predicting the genotype or phenotype of interest by the decoded latent representation; and selecting one or more plant populations or members thereof based on the imputed or predicted genotype or phenotype of interest, wherein the imputed or predicted genotype or phenotype of interest is associated with an agronomic trait.

25. A method of parametrically representing genotypic data obtained from one or more maize populations or sample set to impute or predict a genotype of interest associated with an agronomic trait, the method comprising:

generating a universal latent space representation by encoding discrete or continuous variables derived from genotypic data into latent vectors through a variational autoencoder, wherein the latent space representation is independent of the underlying genotypic data;

decoding the latent space representation by a decoder, thereby imputing or predicting the genotype of interest by the decoded latent representation; and

selecting one or more maize populations or members thereof based on the imputed or predicted genotype or phenotype of interest, wherein the imputed or predicted genotype or phenotype of interest is associated with an agronomic trait.