# US 2019/0244124 A1 · Biosignature Discovery for Substance Use Disorder Using Statistical Learning

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2019/0244124 A1
- Document ID: US20190244124A1
- Status: Published application
- Kind: A1
- Application number: 16/267,360
- Filed: 2019-02-04
- Published: 2019-08-08
- Earliest claimed: 2018-02-02
- Assignees: BioRealm LLC
- Applicants: BioRealm LLC
- Inventors: James W. Baurley; Christopher S. McMahan; Andrew W. Bergen
- Web page: https://patentgrep.com/patent/US20190244124A1

## CPC

- G06N 7/00: Computing arrangements based on specific mathematical models
- G06F 17/18: for evaluating statistical data {, e.g. average values, frequency distributions, probability functions, regression analysis (forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04)}
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 20/00: Machine learning
- G06N 20/20: Ensemble learning
- G16B 20/00: ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
- G16B 40/20: Supervised data analysis
- G06N 5/02: Knowledge representation; Symbolic representation
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G16B: BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY

## Abstract

A method for biomarker discovery for substance use disorders wherein high dimensional data containing a plurality of variables based on a sample size and a number of variables exceeding that sample size are applied to an ensemble of statistical learning models whereby biomarkers of a substance use disorder are identified.

## Claims

### Claim 1

A method for biomarker discovery for substance use disorders, comprising:
a. receiving high dimensional data containing a plurality of variables based on a sample size, wherein the number of variables exceeds the sample size;
b. applying the high dimensional data to an ensemble of statistical learning models, whereby a final biomarker of the substance use disorder is identified.

### Claim 2 (depends on 1)

The method of claim 1, wherein each of the statistical learning models in the ensemble apply a different penalized regression technique for regularization of the high dimensional data.

### Claim 3 (depends on 2)

The method of claim 2, wherein each of the statistical learning models identifies an initial biomarker.

### Claim 4 (depends on 3)

The method of claim 3, wherein the initial biomarkers identified from each of the statistical learning models are filtered by weighted averaging to identify the final biomarker of the substance use disorder.

### Claim 5 (depends on 3)

The method of claim 3, wherein the initial biomarkers from each of the statistical learning models are filtered by voting techniques to identify the final biomarker of the substance use disorder.

### Claim 6 (depends on 2)

The method of claim 2, wherein the ensemble of statistical learning models apply a Bayesian linear regression technique.

### Claim 7 (depends on 6)

The method of claim 6, wherein each of the statistical learning models identifies an initial biomarker.

### Claim 8 (depends on 7)

The method of claim 7, wherein the initial biomarkers from each of the statistical learning models are filtered by weighted averaging to identify the final biomarker of the substance use disorder.

### Claim 9 (depends on 6)

The method of claim 6, wherein the initial biomarkers from each of the statistical learning models are filtered by voting techniques to identify the final biomarker of the substance use disorder.

### Claim 10 (depends on 1)

The method of claim 1, further comprising validating the final biomarker using a k-fold cross validation technique.

### Claim 11 (depends on 1)

The method of claim 1, wherein the high dimensional data is genomic data.

### Claim 12

A method for developing a predictive model for substance use disorders, comprising:
a. receiving high dimensional data that comprises a number of variable that exceeds a sample size;
b. applying a plurality of statistical learning models;
c. presenting a plurality of candidates for initial biomarkers;
d. filtering the plurality of candidates for initial biomarkers to form a final biomarker;
e. presenting the final biomarker;
f. wherein the final biomarker is used in the predictive model.

### Claim 13 (depends on 12)

The method of claim 12, wherein each of the plurality of statistical learning models use penalized regression techniques for regularization of the high dimensional data.

### Claim 14 (depends on 13)

The method of claim 13, wherein each of the plurality of statistical learning models use Bayesian linear regression techniques.

### Claim 15 (depends on 12)

The method of claim 12, wherein the plurality of statistical learning models are stacked into an ensemble.

### Claim 16 (depends on 12)

The method of claim 12, wherein the plurality of candidates for initial biomarkers are filtered using voting techniques.

### Claim 17 (depends on 12)

The method of claim 12, wherein the plurality of candidates for initial biomarkers are filtered using weighted averaging techniques.

### Claim 18 (depends on 12)

The method of claim 12, further comprising validating the final biomarker using a k-fold cross validation technique.

### Claim 19 (depends on 12)

The method of claim 12, wherein the high dimensional data is genomic data.

### Claim 20 (depends on 12)

The method of claim 12, wherein the predictive model is selected from the group consisting of:
a. diagnosing a disease; and
b. identifying a treatment protocol.
