US 10,896,742 B2 · Estimation of phenotypes using DNA, pedigree, and historical data

Abstract

Disclosed are techniques for predicting a trait of an individual and identifying a set of enriched record collections of a genetic community. To predict a trait of an individual, DNA features and non-DNA features of the individual are accessed to generate a feature vector that is inputted into a machine learning model. The machine learning model generates a prediction of the trait. The prediction may be based on an inheritance prediction and/or a community prediction. To identify a set of enriched record collections, individuals belonging to a genetic community are identified and a set of candidate record collections are accessed. A community count and a background count is determined for each candidate record collection. The set of enriched record collections are identified based on a comparison of the community count and the background count. The genetic community may be annotated using the set of enriched record collections.

Claims26

Dependent claims

A computer-implemented method for improving a machine learning model used for trait detection using genetic data, the computer-implemented method comprising:

collecting a set of DNA features of sample individuals;
collecting a set of non-DNA features of the sample individuals;
applying one or more transformations to the set of non-DNA features to generate numerical values representing the non-DNA features;
creating a training set comprising feature vectors that combine the set of DNA features and the set of non-DNA features;
training the machine learning model using the training set;
accessing a target set of DNA features of a target individual;
accessing a target set of non-DNA features of the target individual;
generating a target feature vector that combines the target set of DNA features and the target set of non-DNA features, the target feature vector including a target set of numerical values, at least a first one of the numerical values representing one of the DNA features, at least a second one of the numerical values representing one of the non-DNA features;
and
inputting the target feature vector to the machine learning model to generate a prediction of whether the target individual has a trait.

7 more elements

A computer-implemented method of predicting a trait of a target individual, comprising:

accessing a first dataset representing a family tree of the target individual, the family tree describing relationships among the target individual and related individuals who are related to the target individual;
collecting a set of historical records of the related individuals, the set of historical records comprising at least a first type of features and a second type of features;
applying one or more transformations to the first type of features and the second type of features to generate feature vectors including numerical values that represent the historical records;
inputting the feature vectors into a machine learning model to determine one or more traits of one or more related individuals;
generating, from the first dataset and the one or more traits determined by the machine learning model, a second dataset representing a trait tree, the trait tree comprising nodes representing the target individual and one or more related individuals, each node being connected to at least another node in accordance with the relationships described in the family tree, the nodes comprising (i) a first node representing the target individual, the first node having an unknown value, and (ii) a second node having a known value representing one of the traits of one of the related individuals represented by the second node, the one of traits determined by the machine learning model;
determining the unknown value of the first node, the unknown value determined at least based on an inheritance probability propagated from the known value of the second node along one or more branches of the trait tree, determining the unknown value comprising:
iteratively, by a processor, passing through the trait tree in forwards and backwards directions, and
updating, by the processor, the inheritance probability until the unknown value and the inheritance probability converge;
and
generating a prediction of the trait of the target individual based on a determined value of the first node.

7 more elements

Dependencies are inferred from the claim text.

Estimation of phenotypes using DNA, pedigree, and historical data

Life of this patent

Estimation of phenotypes using DNA, pedigree, and historical data

Life of this patent

Abstract

Claims26