Methods and systems for determining and displaying pedigrees

1. A method, implemented using a computer system that includes one or more processors and system memory, for training a probabilistic relationship model for predicting genetic relationships based on Identity-by-Descent (IBD) data and age data, the method comprising:

a) receiving, by the computer system from a database, IBD data of a plurality of training sets for a plurality of relationships, wherein each training set comprises one or more pairs of individuals, each pair of individuals being genetically related with a same relationship, each training set is associated with a unique relationship, and the IBD data of each training set comprise pairwise IBD data of individuals in the training set;

b) obtaining, by the computer system, age data of the plurality of training sets for the plurality of relationships, wherein the age data of each training set comprise pairwise age data of individuals in the training set; and

c) training, by the computer system, the probabilistic relationship model using the IBD data of the plurality of training sets and the age data of the plurality of training sets, wherein the trained probabilistic relationship model is configured to take as input pairwise IBD data for two test individuals including a length of IBD segments and a number of IBD segments and pairwise age data for the two test individuals and provide as output various likelihoods of various potential relationships for the two test individuals, and wherein each likelihood of each potential relationship for the two individuals is a composite likelihood obtained by combining a likelihood given a length of IBD segment, a likelihood given a number of IBD segments, and a likelihood given an age difference between the two individuals.

2. The method of claim 1 , further comprising:

a) identifying, among the plurality of genetically related individuals, a closest relative of a starting individual using genetic data of the plurality of genetically related individuals;

b) applying pairwise Identity-by-Descent (IBD) data and pairwise age data of the starting individual and the closest relative to a probabilistic relationship model to obtain various likelihoods of various potential relationships between the starting individual and the closest relative;

c) selecting one or more potential relationships between the starting individual and the closest relative that have relationship likelihoods meeting a relationship criterion, and forming a pedigree from each ofthe one or more potential relationships;

d) identifying, among genetically related individuals not included in pedigrees already formed, a closest relative of any individual already in a pedigree;

e) applying pairwise IBD data and pairwise age data of the closest relative and the individual already in the pedigree to the probabilistic relationship model to obtain various likelihoods of various potential relationships between the closest relative and the individual already in the pedigree;

f) selecting one or more potential relationships between the closest relative and the individual already in the pedigree that have relationship likelihoods meeting the relationship criterion, and adding each of the one or more potential relationships to each pedigree already formed to grow each pedigree into one or more growing pedigrees;

g) selecting growing pedigrees that have pedigree likelihoods meeting a pedigree criterion as the pedigrees already formed; and

h) repeating ( 1 d)-( 1 g) one or more times,

wherein ( 1 a)-( 1 h) are performed by the computer system.

3. The method of claim 1 , wherein the pairwise IBD data comprise a length of full IBD segments and/or a length of half IBD segments.

4. The method of claim 1 , wherein the pairwise IBD data comprise a number of full IBD segments and/or a number of half IBD segments.

5. The method of claim 1 , wherein the probabilistic relationship model is a machine-learning model.

6. The method of claim 1 , wherein the probabilistic relationship model models a probability distribution of the pairwise IBD data for each relationship and/or a probability distribution of the pairwise age data for each relationship as a Gaussian distribution, a Poisson distribution, or an exponential distribution.

7. The method of claim 6 , wherein training the probabilistic relationship model comprises using a method-of-moments technique to fit the probability distribution to training data.

8. The method of claim 1 , wherein the plurality of relationships comprise relationships of the 4 th , 5 th , 6 th , 7 th , 8 th , 9 th , 10 th , 11 th , 12 th , 13 th , 14 th or 15 th degree or further.

9. The method of claim 1 , wherein the various relationships comprise relationships of at least 4 meioses on a common-ancestor path.

10. The method of claim 1 , wherein the various relationships comprise relationships having a coefficient of relationship smaller than 0.25.

11. The method of claim 1 , wherein each training set comprises hundreds of individuals.

12. The method of claim 1 , wherein each training set comprises thousands of individuals.

13. The method of claim 1 , wherein each training set comprises tens of thousands of individuals.

14. The method of claim 1 , wherein each training set comprises data of actual individuals.

15. The method of claim 1 , wherein each training set comprises data of simulated individuals.

16. A system comprising one or more processors and one or more computer-readable storage media having stored thereon instructions, said one or more processors being configured to:

a) receive, from a database, Identity-by-Descent (IBD) data of a plurality of training sets for a plurality of relationships, wherein each training set comprises one or more pairs of individuals, each pair of individuals being genetically related with a same relationship, each training set is associated with a unique relationship, and the IBD data of each training set comprise pairwise IBD data of individuals in the training set;

b) obtain age data of the plurality of training sets for the plurality of relationships, wherein the age data of each training set comprise pairwise age data of individuals in the training set; and

c) train a probabilistic relationship model using the IBD data of the plurality of training sets and the age data of the plurality of training sets, wherein the trained probabilistic relationship model is configured to take as input pairwise IBD data for two test individuals including a length of IBD segments and a number of IBD segments and pairwise age data for the two test individuals and provide as output various likelihoods of various potential relationships for the two test individuals, and wherein each likelihood of each potential relationship for the two individuals is a composite likelihood obtained by combining a likelihood given a length of IBD segment, a likelihood given a number of IBD segments, and a likelihood given an age difference between the two individuals.

17. A computer program product comprising one or more computer-readable non-transitory storage media having stored thereon computer-executable instructions that, when executed by one or more processors of a computer system, cause the computer system to:

a) receive, from a database, Identity-by-Descent (IBD) data of a plurality of training sets for a plurality of relationships, wherein each training set comprises one or more pairs of individuals, each pair of individuals being genetically related with a same relationship, each training set is associated with a unique relationship, and the IBD data of each training set comprise pairwise IBD data of individuals in the training set;

b) obtain age data of the plurality of training sets for the plurality of relationships, wherein the age data of each training set comprise pairwise age data of individuals in the training set; and

c) train a probabilistic relationship model using the IBD data of the plurality of training sets and the age data of the plurality of training sets, wherein the trained probabilistic relationship model is configured to take as input pairwise IBD data for two test individuals including a length of IBD segments and a number of IBD segments and pairwise age data for the two test individuals and provide as output various likelihoods of various potential relationships for the two test individuals, and wherein each likelihood of each potential relationship for the two individuals is a composite likelihood obtained by combining a likelihood given a length of IBD segment, a likelihood given a number of IBD segments, and a likelihood given an age difference between the two individuals.