IMPROVED COMPUTER IMPLEMENTED METHOD FOR PREDICTING TRUE AGRONOMICAL VALUE OF A PLANT

• US 20190228839A1
• Filed: 07/23/2015
• Published: 07/25/2019
• Est. Priority Date: 07/23/2015
• Status: Active Grant

• Current Assignee: Limagrain Europe (Groupe Limagrain Holding SA)
• Original Assignee: Limagrain Europe (Groupe Limagrain Holding SA)
• Inventors: HESLOT, Nicolas, CHAUVET, Stephanie, BOYARD, Chloe, FLAMENT, Pascal
• Granted Patent: US 11,430,542 B2
• Time in Patent Office: Days
• CPC Class Codes: C12Q 1/68 involving nucleic acids
• CPC Class Codes: C12Q 1/6895 for plants, fungi or algae
• CPC Class Codes: C12Q 2600/13 Plant traits
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G16B 20/00 ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
• CPC Class Codes: G16B 20/20 Allele or variant detection, e.g. single nucleotide polymorphism [SNP] detection
• CPC Class Codes: G16B 20/40 Population genetics; Linkage disequilibrium
• CPC Class Codes: G16B 40/00 ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
• CPC Class Codes: G16B 40/30 Unsupervised data analysis
• CPC Class Codes: G16B 5/20 Probabilistic models
• CPC Class Codes: G16B 50/20 Heterogeneous data integration

Abstract

A computer implemented method for predicting an agronomical value and a breeding value of a plant belonging to a population, the method includes the steps of: obtaining at least some genotypic data from a subset of lines from the population, obtaining at least some phenotypic data from a subset of lines from the population, providing a statistical model receiving in input the genotypic data and phenotypic data, using the statistical model to output at least an agronomical value estimated for the plant. More particularly, the statistical model is a mixed model combining fixed effects and random effects.

Claims

1. A computer implemented method for predicting an agronomical value and a breeding value of a plant belonging to a population, the method comprising the steps of:

obtaining at least some genotypic data from a subset of lines from the population,

obtaining at least some phenotypic data from a subset of lines from the population,

providing a statistical model receiving in input said genotypic data and phenotypic data,

using said statistical model to output at least an agronomical value estimated for said plant,

wherein said statistical model is a mixed model combining fixed effects and random effects.

2. The method of claim 1, wherein the mixed model is used to output further a predicted breeding value estimated for said plant.

3. The method according to claim 1, wherein said genotypic data comprise genetic markers data.

4. The method according to claim 1, wherein said mixed model further receives, in input, phenotypic data of said plant.

5. The method of claim 4, wherein said phenotypic data comprise field trial information data.

6. The method according to claim 1, wherein said mixed model comprises a Genomic Best Linear Unbiased Prediction (GBLUP).

7. The method according to claim 1, wherein said random effects include at least non-additive genetic effects,

8. The method according to claim 7, wherein a Gaussian Kernel or others marker-based kinships are used to estimate a covariance between plant individuals and capture non-additive genetic effects.

9. The method according to claim 1, comprising further a step of estimating a genetic dominance effect and/or a non-additive effect based on genotypic data to capture heterosis, so as to predict hybrid crops performance.

10. The method according to claim 1, wherein said mixed model further receives, in input, phenotypic data of said plant, and wherein said mixed model analyses simultaneously several phenotype traits with a covariance estimated between traits and/or estimated for a same trait being observed in several different environments.

11. The method according to claim 1, wherein said mixed model uses at least one additional fixed effect covariate to model an effect of genotypic data on phenotypic data of said plant.

12. The method according to claim 1, wherein said mixed model has heterogeneous residuals.

13. The method according claim 1, comprising a set of tests on procedure to select the most predictive model.

14. The method according to claim 1, comprising further a preliminary step of filtering at least said genotypic data so as to remove at least samples having incomplete data.

15. The method according to claim 1, wherein said mixed model further receives, in input, phenotypic data of said plant, and wherein automatic detection and removal of outliers phenotypic observations and/or genotypic individuals are performed.

16. A computer program comprising instructions to perform the method according to claim 1, when such instructions are executed by a processor.

17. A device for predicting an agronomical value of a plant, comprising a circuit including at least:

an input for receiving at least some genotypic data and some phenotypic data.

a processor for executing the method of the invention according to claim 1,

and an output to deliver at least the agronomical value of said plant.