Feature selection for efficient epistasis modeling for phenotype prediction

Patent metadata

Field	Value
Publication number	US-10102333-B2
Application number	US-201313745914-A
Country	US
Kind code	B2
Filing date	Jan 21, 2013
Priority date	Jan 21, 2013
Publication date	Oct 16, 2018
Grant date	Oct 16, 2018

Abstract

Various embodiments select markers for modeling epistasis effects. In one embodiment, a processor receives a set of genetic markers and a phenotype. A relevance score is determined with respect to the phenotype for each of the set of genetic markers. A threshold is set based on the relevance score of a genetic marker with a highest relevancy score. A relevance score is determined for at least one genetic marker in the set of genetic markers for at least one interaction between the at least one genetic marker and at least one other genetic marker in the set of genetic markers. The at least one interaction is added to a top-k feature set based on the relevance score of the at least one interaction satisfying the threshold.

First claim

What is claimed is:

1. A method, by a distributed computing system comprising a plurality of information processing systems, for reducing search time of features in sample data and reducing computation time when training a model for generating specialized data based on the features, the method comprising:
electronically communicating, by a first feature selection circuit of a first information processing system in the distributed computing system, with at least second information processing system in the distributed computing system that is external to the first information processing system;
obtaining, by the first feature selection circuit based on electronically communicating with the second information processing system, a set of genetic data comprising a first set of genetic markers and a phenotype, and a second set of genetic markers, where the second set of genetic markers is the first set of genetic markers absent the phenotype;
training, by a second feature selection circuit of a third information processing system in the distributed computing system, an epistasis effect model based on the set of genetic markers and the phenotype;
reducing, by the first feature selection circuit and the second feature selection circuit, computation time of the distributed processing system during the training of the epistasis effect model and further reducing feature selection time of the epistasis effect model, wherein reducing the computation time and the feature selection time comprises:
determining, by the first feature selection circuit for each of the first set of genetic markers, a relevance score with respect to the phenotype according to I(x j training ;c training ), where I is mutual information between a given genetic marker x j and a phenotype c, where mutual information I between two variables x and y is defined, based on their joint marginal probabilities p(x) and p(y) and probabilistic distribution p(x, y), as:

I
⁡

(

x
,
y

)

=

∑

i
,
j

⁢

p
⁡

(

x
i

,

y
i

)

⁢
log
⁢

⁢

p
⁡

(

x
i

,

y
i

)

p
⁡

(

x
i

)

⁢

p
⁡

(

y
i

)

;

setting, by the first feature selection circuit, a threshold to the relevance score of a genetic marker in the first set of genetic markers with a highest relevancy score;
determining, by the first feature selection circuit for at least one individual genetic marker in the first set of genetic markers having a relevance score satisfying the threshold, a relevance score for at least one interaction between the at least one individual genetic marker and at least one other individual genetic marker in the first set of genetic markers;
adding, by the first feature selection circuit, the at least one interaction to a top-k feature set based on the relevance score of the at least one interaction satisfying the threshold, wherein the top-k feature set comprises one or more markers and one or more interactions, and wherein each of the one or more genetic markers and each of the one or more interactions comprises a top-k relevance score;
identifying, by the first feature selection circuit, a subset of the top-k feature set based on both the first set and second set of genetic markers, wherein each feature in the subset of the top-k feature set maximizes a relevancy with the phenotype and minimizes a redundancy with respect to other selected features; and
transmitting, by the first feature selection circuit, at least the subset of the top-k feature set to at least the third information processing system in the distributed computing system, and programming the second feature selection of the third information processing system to perform one or more operations comprising
training, by the second feature selection circuit of the third information processing system, an epistasis effect model utilizing at least the subset of the top-k feature set, wherein the epistasis effect model predicts phenotypes for genetic markers;

storing, by the second feature selection circuit, the trained epistasis effect model in memory; and
executing, by the third information processing system, the trained epistasis effect model, wherein executing the trained epistasis effect model comprises
electronically obtaining, by the second feature selection circuit, a new set of a set of genetic markers that is associated phenotype data;
inputting, by the second feature selection circuit, the new set of genetic markers into the trained epistasis effect model; and

outputting, by the second feature selection circuit based on the inputting, a phenotype for the new set of genetic markers, wherein the phenotype that was outputted was not made available to the feature selection circuit as part of the new set of genetic markers.

2. The computer implemented method of claim 1 , wherein the relevance score determined for each of the first set of genetic markers is based on mutual information between the each of the first set of genetic markers and the phenotype.

3. The computer implemented method of claim 1 , wherein the relevance score determined for the at least one interaction is based on mutual information between the at le

Assignees

• IBM

Inventors

• Haws David
• He Dan
• Parida Laxmi P

Classifications

• G06F19/18

  Physics · mapped topic

• G06F19/12 Primary

  Physics · mapped topic

• G06F19/24

  Physics · mapped topic

• G16B5/20 Primary

  Probabilistic models · CPC title

• G16B40/00

  ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding · CPC title