Granted Patent B2

US 12,639,342 · App. 18/924,937 · Granted May 26, 2026

Inventors: Milos Pavlovic (Sandy, UT); Ross Eugene Curtis (Cedar Hills, UT); Yong Wang (Foster City, CA); Luong Ruiz (San Francisco, CA)

Assignee: Ancestry.com DNA, LLC

G06F16/288G06F16/285

Abstract

Claims (96)

1 . A computer-implemented method, comprising:

receiving a first genetic dataset associated with a target individual;

receiving a second genetic dataset associated with a target relative, wherein the target relative is a potential relative to the target individual;

determining a total length of genetic segments that are matched between the first genetic dataset and the second genetic dataset;

determining that the total length falls within a predetermined range of a degree of separation;

identifying a plurality of genetic matches, each genetic match being an individual having genetic data that matches the first genetic dataset, the second genetic dataset, or both;

extracting features from the genetic data of the plurality of genetic matches; and

inputting the extracted features into a machine learning model to determine a familial relationship between the target individual and the target relative, the machine learning model specifically trained for the predetermined range of the degree of separation.

2 . The computer-implemented method of claim 1 , wherein identifying the plurality of genetic matches comprises:

determining a data classification link between the target individual and the target relative based on a degree of match between the first genetic dataset and the second genetic dataset; and

identifying the plurality of genetic matches based on the determined data classification link.

3 . The computer-implemented method of claim 1 , wherein extracting the features from the plurality of genetic matches comprises:

determining:

a first length of matched-data-string segments shared by the target individual and the each genetic match,

a second length of matched-data-string segments shared by the target relative and the each genetic match, and

a length difference of matched-data-string segments based on the first length of matched-data-string segments and the second length of matched- data-string segments;

determining:

a first number of matched-data-string segments shared by the target individual and the each genetic match,

a second number of matched-data-string segments shared by the target relative and the each genetic match, and

a number difference of matched-data-string segments based on the first number of matched-data-string segments and the second number of matched-data-string segments;

determining:

a first number of shared classification-data segments between the target individual and the each genetic match,

a second number of shared classification-data segments between the target relative and the each genetic match, and

a number difference of shared classification-data segments based on the first number of shared classification-data segments and the second number of shared classification-data segments; and

extracting the features from the length difference of matched-data-string segments, the number difference of matched-data-string segments, or the number difference of shared classification-data segments.

4 . The computer-implemented method of claim 3 , wherein the features comprise:

a mean, a median, a standard deviation, a maximum or a minimum of the length differences of matched-data-string segments;

a mean, a median, a standard deviation, a maximum or a minimum of the number differences of matched-data-string segments; or

a mean, a median, a standard deviation, a maximum or a minimum of the number difference of shared classification-data segments.

5 . The computer-implemented method of claim 1 , wherein the machine learning model comprises a regression model, a random forest classifier, a support vector machine, a neural network, or a model trained by an unsupervised approach.

6 . The computer-implemented method of claim 1 , further comprising connecting the target individual and the target relative in a data tree based on the familial relationship.

7 . The computer-implemented method of claim 1 , wherein the familial relationship between the target individual and the target relative is an M3 relationship.

8 . The computer-implemented method of claim 1 , further comprising connecting the target individual and the target relative in a data tree based on the familial relationship.

9 . A system comprising:

one or more processors; and

memory configured to store instructions, wherein the instructions, when executed by the one or more processors, cause the one or more processors to perform steps comprising:

receiving a first genetic dataset associated with a target individual;

receiving a second genetic dataset associated with a target relative, wherein the target relative is a potential relative to the target individual;

determining a total length of genetic segments that are matched between the first genetic dataset and the second genetic dataset;

determining that the total length falls within a predetermined range of a degree of separation;

identifying a plurality of genetic matches, each genetic match being an individual having genetic data that matches the first genetic dataset, the second genetic dataset, or both;

extracting features from the genetic data of the plurality of genetic matches; and

inputting the extracted features into a machine learning model to determine a familial relationship between the target individual and the target relative, the machine learning model specifically trained for the predetermined range of the degree of separation.

10 . The system of claim 9 , wherein identifying the plurality of genetic matches comprises:

determining a data classification link between the target individual and the target relative based on a degree of match between the first genetic dataset and the second genetic dataset; and

identifying the plurality of genetic matches based on the determined data classification link.

11 . The system of claim 9 , wherein extracting the features from the plurality of genetic matches comprises:

determining:

a first length of matched-data-string segments shared by the target individual and the each genetic match,

a second length of matched-data-string segments shared by the target relative and the each genetic match, and

a length difference of matched-data-string segments based on the first length of matched-data-string segments and the second length of matched- data-string segments;

determining:

a first number of matched-data-string segments shared by the target individual and the each genetic match,

a second number of matched-data-string segments shared by the target relative and the each genetic match, and

a number difference of matched-data-string segments based on the first number of matched-data-string segments and the second number of matched-data-string segments;

determining:

a first number of shared classification-data segments between the target individual and the each genetic match,

a second number of shared classification-data segments between the target relative and the each genetic match, and

a number difference of shared classification-data segments based on the first number of shared classification-data segments and the second number of shared classification-data segments; and

extracting the features from the length difference of matched-data-string segments, the number difference of matched-data-string segments, or the number difference of shared classification-data segments.

12 . The system of claim 11 , wherein the features comprise:

a mean, a median, a standard deviation, a maximum or a minimum of the length differences of matched-data-string segments;

a mean, a median, a standard deviation, a maximum or a minimum of the number differences of matched-data-string segments; or

a mean, a median, a standard deviation, a maximum or a minimum of the number difference of shared classification-data segments.

13 . The system of claim 9 , wherein the machine learning model comprises a regression model, a random forest classifier, a support vector machine, a neural network, or a model trained by an unsupervised approach.

14 . The system of claim 9 , wherein the familial relationship between the target individual and the target relative is an M3 relationship.

15 . A non-transitory computer readable medium for storing computer code comprising instructions, when executed by one or more computer processors, causing one or more computer processors to perform steps comprising:

receiving a first genetic dataset associated with a target individual;

receiving a second genetic dataset associated with a target relative, wherein the target relative is a potential relative to the target individual;

determining a total length of genetic segments that are matched between the first genetic dataset and the second genetic dataset;

determining that the total length falls within a predetermined range of a degree of separation;

identifying a plurality of genetic matches, each genetic match being an individual having genetic data that matches the first genetic dataset, the second genetic dataset, or both;

extracting features from the genetic data of the plurality of genetic matches; and

inputting the extracted features into a machine learning model to determine a familial relationship between the target individual and the target relative, the machine learning model specifically trained for the predetermined range of the degree of separation.

16 . The non-transitory computer readable medium of claim 15 , wherein identifying the plurality of genetic matches comprises:

determining a data classification link between the target individual and the target relative based on a degree of match between the first genetic dataset and the second genetic dataset; and

identifying the plurality of genetic matches based on the determined data classification link.

17 . The non-transitory computer readable medium of claim 15 , wherein extracting the features from the plurality of genetic matches comprises:

determining:

a first length of matched-data-string segments shared by the target individual and the each genetic match,

a second length of matched-data-string segments shared by the target relative and the each genetic match, and

a length difference of matched-data-string segments based on the first length of matched-data-string segments and the second length of matched- data-string segments;

determining:

a first number of matched-data-string segments shared by the target individual and the each genetic match,

a second number of matched-data-string segments shared by the target relative and the each genetic match, and

a number difference of matched-data-string segments based on the first number of matched-data-string segments and the second number of matched-data-string segments;

determining:

a first number of shared classification-data segments between the target individual and the each genetic match,

a second number of shared classification-data segments between the target relative and the each genetic match, and

a number difference of shared classification-data segments based on the first number of shared classification-data segments and the second number of shared classification-data segments; and

extracting the features from the length difference of matched-data-string segments, the number difference of matched-data-string segments, or the number difference of shared classification-data segments.

18 . The non-transitory computer readable medium of claim 17 , wherein the features comprise:

a mean, a median, a standard deviation, a maximum or a minimum of the length differences of matched-data-string segments;

a mean, a median, a standard deviation, a maximum or a minimum of the number differences of matched-data-string segments; or a mean, a median, a standard deviation, a maximum or a minimum of the number difference of shared classification-data segments.

19 . The non-transitory computer readable medium of claim 15 , wherein the machine learning model comprises a regression model, a random forest classifier, a support vector machine, a neural network, or a model trained by an unsupervised approach.

20 . The non-transitory computer readable medium of claim 15 , wherein the familial relationship between the target individual and the target relative is an M3 relationship.

Assignments (3)

Continuity (2)

References Cited (142)