Methods for identifying crosses for use in plant breeding

Patent metadata

Field	Value
Publication number	US-10327400-B2
Application number	US-201715618023-A
Country	US
Kind code	B2
Filing date	Jun 8, 2017
Priority date	Jun 8, 2016
Publication date	Jun 25, 2019
Grant date	Jun 25, 2019

Abstract

Exemplary methods for use in identifying crosses for use in plant breeding are disclosed. One exemplary method includes selecting a subgroup of potential crosses, based on thresholds associated with population prediction scores for the set of potential crosses. The exemplary method further includes selecting multiple target crosses from the subgroup of potential crosses based on a genetic relatedness of the parents in the subgroup of potential crosses, filtering the target crosses based on a rule (or rules) defining a threshold (or thresholds) for at least one characteristic and/or trait, selecting ones of the filtered target crosses based on risk associated with the selected one of the filtered target crosses, and directing the selected ones of the filtered target crosses into a breeding pipeline, thereby providing crosses to the breeding pipeline based, at least in part, on commercial success of parents included in the selected ones of the filtered crosses.

First claim

What is claimed is:

1. A method for use in identifying crosses for use in plant breeding, the method comprising:
accessing a data structure representative of multiple parents;
identifying a set of potential crosses, each potential cross in the set of potential crosses including at least two of the multiple parents included in the data structure;
selecting, by at least one computing device, a subgroup of potential crosses, from the set of potential crosses, based on one or more thresholds associated with population prediction scores for the set of potential crosses, each population prediction score associated with a prediction of commercial success for the associated potential cross within the set of potential crosses;
selecting, by the at least one computing device, multiple target crosses from the subgroup of potential crosses based on a genetic relatedness of the parents in the subgroup of potential crosses;
filtering, by the at least one computing device, the target crosses based on at least one rule, the at least one rule defining at least one threshold for at least one characteristic and/or trait of at least one of: the multiple target crosses, one of the multiple parents included in the target crosses, and a parental line of the target crosses;
selecting, by the at least one computing device, ones of the filtered target crosses based on risk associated with the selected one of the filtered target crosses; and
including a plant in a growing space of a breeding pipeline, the plant derived from at least one of the selected ones of the filtered target crosses.

2. The method of claim 1 , further comprising generating, by the at least one computing device, the population prediction scores for each potential cross within the set of potential crosses.

3. The method of claim 2 , wherein generating the population prediction scores includes generating, by the at least one computing device, the population prediction scores based on the following algorithm:
p ( s i |x i ,D )=Σ m=1 M p ( s i |x i ,m,D ) p ( m|D ).

4. The method of claim 1 , wherein selecting the multiple target crosses from the subgroup of potential crosses based on the genetic relatedness includes:
clustering, by the at least one computing device, the parents of the potential crosses included in the subgroup, based on the genetic relatedness of the parents; and
selecting the multiple target crosses based on a relatedness threshold associated with the clustered parents of the target crosses.

5. The method of claim 4 , wherein clustering the parents includes spectral clustering, by the at least one computing device, of the parents of the potential crosses included in the subgroup; and
wherein selecting the multiple target crosses includes:
combining, by the at least one computing device, a cluster score associated with at least one parent of one of the potential crosses included in the subgroup and a cluster score associated with said one of the potential crosses; and
selecting the multiple target crosses based on a comparison of the combined cluster scores to the relatedness threshold.

6. The method of claim 1 , wherein the at least one rule is associated with at least one of stalk lodging, root lodging, Goss Wilt, parental similarity, and a difference between expected relative maturity (ERM) between the two parents.

7. The method of claim 1 , wherein selecting ones of the filtered target crosses based on risk associated therewith includes determining, by the at least one computing device, risks associated with the potential crosses based on a quadratic algorithm dependent on a risk variable, a diversity variable, and a performance variable of the crosses.

8. The method of claim 7 , wherein determining the risks includes determining the risks, by the at least one computing device, based on:
x OPT =arg maxλ perf ( c T x+x T Px )−½(λ risk x T Rx+λ div x T Sx )

subject to Σx i =1, x i ≥0∀i and
Σ x i ϵF x i ≥0.4 and Σ x i ϵM x i ≥0.4.

9. The method of claim 7 , wherein the risk variable of the quadratic algorithm associated with the ones of the filtered target crosses accounts for risk associated with multiple characteristics and/or traits of the cross.

10. A system for use in identifying crosses for use in plant breeding, the system comprising:
a breeding pipeline;
a crosses data structure including multiple parents available for use in crosses in plant breeding, and a set of potential crosses, each potential cross in the set of potential crosses including at least two of the multiple parents;
a computing device coupled in communication with the data structure and configured to:
select a subgroup of potential crosses from the set of potential crosses, in the data structure, based on one or more thresholds associated with population prediction scores for the potential crosses;
select multiple target crosses, from the subgroup of potential crosses, based on genetic relatedness of the parents of the subgroup of potential crosses;
filter the target crosses based on at least one rule, the at least one rule defining at least one threshold for at least one characteristic and/or trait of the target crosses, of one of the multiple parents included in the target crosses, and/or of a parental line of the target crosses;
select ones of the filtered target crosses based on a risk associated therewith; and
direct the selected ones of the filtered target crosses into a breeding pipeline, thereby providing crosses to the breeding pipeline based, at least in part, on commercial success of parents to the selected ones of the filtered crosses;
wherein the breeding pipeline includes a growing space and a plant derived from at least one of the selected ones of the filtered target crosses planted in the growing space, after the selected ones of the filtered target crosses are directed into the breeding pipeline.

11. The system of claim 10 , wherein the computing device is further configured to:
intermittently generate the population prediction score for each of the potential crosses in the subgroup of potential crosses; and
store the population prediction scores in the crosses data structure.

12. The system of claim 10 , wherein the computing device is configured to cluster the parents of the potential crosses included in the subgroup, based on the relatedness of the parents of the potential crosses, and then to select the multiple target crosses from the subgroup of potential crosses based on a relatedness threshold associated with the clustered parents of the target crosses, thereby selecting the multiple target crosses based on genetic relatedness of the parents of the subgroup of potential crosses.

13. The system of claim 10 , wherein the computing device is further configured to identify, based on a user input, the set of potential crosses.

14. The system of claim 10 , wherein the computing device is further configured to generate the population prediction score, for each potential cross, based on:
p ( s i |x i ,D )=Σ m=1 M p ( s i |x i ,m,D ) p ( m|D ).

15. The system of claim 10 , wherein, in order to select the multiple target crosses from the subgroup based on genetic relatedness, the computing device is configured to cluster the parents of the target crosses included in the subgroup based on the relatedness of the parents and then to select the multiple target crosses based on a relatedness threshold associated with the clustering of the parents.

16. The system of claim 15 , wherein, in order to cluster the parents, the comp

Assignees

• Monsanto Technology Llc

Inventors

• Chavali Srinivas Phani Kumar
• Dasgupta Sambarta
• Polavarapu Nalini

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06N20/20 Primary

  Ensemble learning · CPC title

• G06N20/00

  Machine learning · CPC title

• G06N7/005

  Physics · mapped topic

• A01H1/04 Primary

  Processes of selection {involving genotypic or phenotypic markers; Methods of using phenotypic markers for selection} · CPC title