# US 2024/0203593 A1 · Method of discovering disease-relevant microbial consortia using machine-learning model

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

- Number: US 2024/0203593 A1
- Document ID: US20240203593A1
- Status: Published application
- Kind: A1
- Application number: 18/537,992
- Filed: 2023-12-13
- Published: 2024-06-20
- Earliest claimed: 2022-12-14
- Assignees: Immunobiome Inc.
- Applicants: Immunobiome Inc.
- Inventors: Seung Yeon CHOI; In Hae KIM; Jung Hwan PARK; Ju Hun LEE
- Web page: https://patentgrep.com/patent/US20240203593A1

## CPC

- G16B 10/00: ICT specially adapted for evolutionary bioinformatics, e.g. phylogenetic tree construction or analysis
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- G16B 20/00: ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
- G16B 40/20: Supervised data analysis
- G16B 40/30: Unsupervised data analysis
- G16B: BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

The present invention aims to discover a microbial cluster that can be used when developing effective probiotics for a specific disease. Specifically, the invention includes (1) discovering a candidate microbial cluster data related to specific disease, which is collected by preprocessing a gut microbiota data (a taxonomy abundance table) and (2) training the machine learning model using the collected candidate microbial cluster data, and selecting a model with the highest predictive performance to discover a disease-microbial consortium.

## Claims

### Claim 1

A method of drawing microbial cluster data using a machine learning model by computing device, comprising:
(1) a process of collecting gut microbiota data to draw a candidate microbial cluster data from the collected data; and
(2) a process of drawing a disease relevant microbial cluster data from the candidate microbial cluster data.

### Claim 2 (depends on 1)

The method according to claim 1,
wherein the process of (1) comprises:
a process of calculating pairwise taxonomic similarity of the gut microbiota data;
a process of clustering the gut microbiota data to generate primary cluster data based on the pairwise taxonomic similarity; and
a process of drawing the candidate microbial cluster data through quality control.

### Claim 3 (depends on 2)

The method according to claim 2,
wherein the clustering is conducted using an unsupervised learning algorithm.

### Claim 4 (depends on 3)

The method according to claim 3,
wherein the unsupervised learning algorithm includes one or more of hierarchical clustering, K-means clustering, and Gaussian mixture model.

### Claim 5 (depends on 2)

The method according to claim 2,
wherein the quality control is to filter out cluster data including either one or two microbes, or those containing more than half of the total microbes in the gut microbiota data.

### Claim 6 (depends on 1)

The method according to claim 1,
wherein the process of (2) comprises:
a process of splitting the candidate microbial cluster data into training set and test set;
a process of training a machine learning model with the training set;
a process of identifying an algorithm with the highest median predictive performance and selecting a machine learning model with the highest predictive performance among from machine learning models using the algorithm; and
a process of drawing the disease relevant microbial cluster data using the selected machine learning model.

### Claim 7 (depends on 6)

The method of claim 6,
wherein the training set and the test set are split using Monte Carlo Random sampling.

### Claim 8 (depends on 6)

The method of claim 6,
wherein the algorithm is a supervised learning algorithm.

### Claim 9 (depends on 8)

The method of claim 8,
wherein the supervised learning algorithm includes one or more of the Logistic regression, Naïve Bayes, Random Forest, and SVM(Support vector machines).

### Claim 10 (depends on 6)

The method of claim 6,
wherein the disease relevant microbial cluster data is drawn as a cluster data with the highest feature importance among from the candidate microbial cluster data used in the selected machine learning model.

### Claim 11

A device of drawing microbial cluster data using a machine learning model by computing device, comprising:
a collecting unit configured to collect gut microbiota data;
a candidate consortium drawing unit configured to draw a candidate microbial cluster data from the collected data; and
a disease-relevant consortium drawing unit configured to draw a disease relevant microbial cluster data from the candidate microbial cluster data.

### Claim 12 (depends on 11)

The device of claim 11,
wherein the candidate consortium drawing unit is further configured to:
calculate pairwise taxonomic similarity of the gut microbiota data;
cluster the gut microbiota data to generate primary cluster data based on the pairwise taxonomic similarity; and
draw the candidate microbial cluster data through quality control.

### Claim 13 (depends on 12)

The device of claim 12,
wherein the clustering is conducted using an unsupervised learning algorithm.

### Claim 14

The device of 13,
wherein the unsupervised learning algorithm includes one or more of hierarchical clustering, K-means clustering, and Gaussian mixture model.

### Claim 15 (depends on 12)

The device of claim 12,
wherein the quality control is to filter out a cluster data including either one or two microbes, or those containing more than half of the total microbes in the gut microbiota data.

### Claim 16 (depends on 11)

The device of claim 11,
wherein the disease-relevant consortium drawing unit is further configure to:
split the candidate microbial cluster data into training set and test set;
train a machine learning model using an algorithm with the training set;
identify the algorithm with the highest median predictive performance and select a machine learning model with the highest predictive performance among from machine learning models using the algorithm; and
draw the disease relevant microbial cluster data using the selected machine learning model.

### Claim 17 (depends on 16)

The device of claim 16,
wherein the training set and the test set are split using Monte Carlo Random sampling.

### Claim 18 (depends on 16)

The device of claim 16,
wherein the algorithm is a supervised learning algorithm.

### Claim 19 (depends on 18)

The device of claim 18,
wherein the supervised learning algorithm includes one or more of the Logistic regression, Naïve Bayes, Random Forest, and SVM (Support vector machines).

### Claim 20 (depends on 16)

The device of claim 16,
wherein the disease relevant microbial cluster data is drawn as a cluster data with the highest feature importance among from the candidate microbial cluster data used in the selected machine learning model.
