# US 2026/0301973 A1 · System and method for identifying pathogen groups of higher virulence by genomic features and for categorizing the groups in accordance with public risks associated therewith

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

- Number: US 2026/0301973 A1
- Document ID: US20260301973A1
- Status: Published application
- Kind: A1
- Application number: 19/479,332
- Filed: 2024-06-19
- Published: 2026-10-01
- Earliest claimed: 2023-06-19
- Assignees: EpiX Analytics LLC
- Applicants: EpiX Analytics LLC
- Inventors: Francisco Javier Zagmutt-Vergara; Jane Gurnick Pouzou; Daniel Davies Taylor; Solenne Francoise Helene Costard; Huybert Groenendaal; Régis Claude Pouillot; Gavin John Fenske
- Web page: https://patentgrep.com/patent/US20260301973A1

## CPC

- G16H 50/80: for detecting, monitoring or modelling epidemics or pandemics, e.g. flu
- G16B 20/00: ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
- G16B 20/20: Allele or variant detection, e.g. single nucleotide polymorphism [SNP] detection
- G16B 40/20: Supervised data analysis
- G16B 40/30: Unsupervised data analysis
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16B: BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY

## Abstract

Certain pathogens are known to cause infection in humans or other living hosts. However, harmful pathogens may have thousands of distinct strains or serovars, only a fraction of which may cause a severe or life-threatening infection. Timely identification of strains which are most likely to cause serious infections is problematic. A system and a method are disclosed which provide a diagnostic protocol for identifying virulence groups in one or more pathogens of interest based on genomic features and/or identifying pathogenic markers of higher virulence in hazardous pathogens such as Salmonella enterica which may be used in combination with pathogen surveillance data to optimize the identification of pathogens of public risk and health concern so that the most effective interventions may be selected and applied in a timely manner.

## Claims

### Claim 1

A method for identifying and grouping isolates of hazardous pathogens by genomic features of higher virulence and associated public health risk, the method comprising the steps of:
identifying the public health risk;
identifying isolates of one or more hazardous pathogens of interest associated with the public health risk;
harvesting or collecting a first dataset including genomic data of each of a plurality of genetic isolates from at least one biological or environmental sample and, when available, associated metadata that describes further information about each of the plurality of genetic isolates;
inputting the first dataset into a bioinformatics process; and
executing the bioinformatics process to arrange, analyze, interpret, and catalog the plurality of genetic isolates into virulence groups based upon their genetic virulence relatedness.

### Claim 2 (depends on 1)

The method of claim 1 further including the step of executing a prioritization process to identify subgroups of highest importance amongst isolates of hazardous pathogens grouped by genomic features of higher virulence and associated public health risk thereof, based on genetic or phenotypic virulence criteria.

### Claim 3 (depends on 1)

The method of claim 1 further including the step of executing a validation process to compare the epidemiological outcome data with the genomic data collected, analyzed, interpreted, and cataloged using bioinformatics processes, whereby the grouping of the genomic data is verified.

### Claim 4 (depends on 3)

The method of claim 3 further including the step of executing a prioritization process to identify subgroups of higher importance based on genetic or phenotypic virulence criteria.

### Claim 5

A method for identifying and grouping pathogenic isolates of hazardous pathogens by genomic features of higher virulence and associated public health risk thereof, and for subsequently identifying pathogenic biomarkers to optimize the detection of hazardous pathogens of higher virulence, the method comprising the steps of:
identifying the public health risk;
identifying isolates of one or more hazardous pathogens of interest associated with the public health risk;
harvesting or collecting a first dataset including genomic data of each of a plurality of genetic isolates from at least one biological or environmental sample and, when available, associated metadata that describes further information about each of the plurality of genetic isolates;
inputting the first dataset into a bioinformatics process;
executing the bioinformatics process to arrange, analyze, interpret, and catalog the plurality of genetic isolates into virulence groups based upon their genetic virulence relatedness; and
executing an optimization process whereby optimal biomarker targets are identified to optimize the identification of higher virulence pathogens.

### Claim 6 (depends on 5)

The method of claim 5 further including the step of executing a prioritization process prior to the step of executing the optimization process to identify subgroups of highest importance amongst isolates of hazardous pathogens grouped by genomic features of higher virulence and associated public health risk thereof, based on genetic or phenotypic virulence criteria.

### Claim 7 (depends on 5)

The method of claim 5 further including the step of executing a validation process prior to the step of executing the optimization process to compare the epidemiological outcome data with the genomic data collected, analyzed, interpreted, and cataloged using biometrics processes, whereby the grouping of the genomic data is verified.

### Claim 8 (depends on 7)

The method of claim 7 further including the steps of executing a prioritization process prior to the step of executing the optimization process, to identify subgroups of higher importance amongst isolates of hazardous pathogens grouped by genomic features of higher virulence and associated public health risk thereof, based on genetic or phenotypic virulence criteria.

### Claim 9

A method for identifying pathogenic biomarkers in isolates of hazardous pathogens grouped by genomic features of higher virulence and associated public health risk thereof, the method comprising the steps of:
identifying the public health risk;
identifying a dataset with isolates of one or more hazardous pathogens of interest associated with the public health risk catalogued into virulence groups based upon their genetic virulence relatedness; and
executing an optimization process whereby optimal biomarker targets are identified to optimize the identification of higher virulence pathogens.

### Claim 10 (depends on 9)

The method of claim 9 further including the step of executing a prioritization process prior to the step of executing the optimization process to identify subgroups of higher importance amongst isolates of hazardous pathogens grouped by genomic features of higher virulence and associated public health risk thereof, based on genetic or phenotypic virulence criteria.

### Claim 11 (depends on 1)

The method of claim 1 wherein the bioinformatics process comprises the steps of:
compiling a first pathogen database based on the plurality of genetic isolates and metadata of interest;
generating a user-defined subset of the plurality genetic isolates and metadata stored in the first pathogen database by defining one or more biological or environmental samples of interest (e.g., a specific commodity or sample type);
applying quality control measures to the user-defined subset of the plurality genetic isolates and metadata for compiling a first quality controlled dataset of selected quality controlled genetic isolates and metadata of interest;
compiling a final quality controlled genetic isolate database by adding a predicted phenotype to each genetic isolate of the pathogen in the first quality controlled database;
selecting at least one reference genome of interest from a genetic isolate in the first quality control database;
developing a model training file for training gene predictions based open reading frames (ORF) of the at least one reference genome of the pathogen of interest;
obtain a reference proteome for the pathogen of interest from a genetic isolate in the first quality control database
compiling a combined pathogen database based on genetic elements and the reference proteome or genome of interest;
compiling a first matrix based upon the final quality controlled database, the at least one reference genome of interest, and the combined pathogen database; and
compiling a second matrix of annotated genetic elements of interest for each isolate.

### Claim 12 (depends on 11)

The method of claim 11 wherein the step of executing the bioinformatics process further includes the steps of:
applying one or more machine learning algorithms to the first dataset to determine the relatedness of a plurality of genetic factors, including genetic virulence, among-the plurality of genetic isolates;
setting a number for user specified groups wherein the number of the user-specified groups is at least two; and
arranging the plurality of genetic isolates of the first dataset into the number of user-specified groups using hierarchical clustering based on the relatedness of the plurality of genetic factors.

### Claim 13 (depends on 9)

The method of claim 9 wherein the optimization process comprises the steps of:
using statistical methods to define a list with a reduced number of genomic features based on genetic or phenotypic groups;
setting classification targets such as maximizing the probability of correctly identifying pathogens of a higher virulence group; and
using algorithms such as constrained optimization or other algorithms to identify the subset of genetic elements that satisfy the classification targets.

### Claim 14 (depends on 3)

The method of claim 3 wherein the validation process includes a first set of steps comprising the steps of:
a first step of accessing a dataset which includes a first subset containing confirmed cases of a plurality of isolates of interest extracted from an epidemiological patient outcome dataset related to the isolates of interest from public health agencies or other healthcare datasets such as the Center for Disease Control (“CDC”), and/or private entities, such as insurance or healthcare providers, including information related to at least one set of isolates from at least one human illness of the pathogen of interest in all user-defined subgroups and the associated outcomes from the at least one human illness or infection as some or all of which have been assigned to one or more user-specified groups by the bioinformatics process based on the virulence genes or genetic elements found in each isolate of the pathogen of interest, and a second subset containing confirmed cases of a plurality of isolates of interest, some of which, if any, that have not been assigned to a user-specified group;
a second step of separating and organizing the first subset of confirmed cases accessed in the first step into a first set of isolates associated with patient data related to confirmed cases of human illness or infection that have been assigned to at least one of the user-specified groups in the bioinformatics process;
a third step of organizing the second subset of the dataset of confirmed cases accessed in the first step that was not assigned to at least one of the user-specified groups;
a fourth step of categorizing the confirmed cases in the third step into two categories wherein a first category includes confirmed cases with recorded subpopulation information such as strain or variant information and no associated or identified genetic sequence data;
a fifth step of associating the first category of confirmed cases without associated or identified genetic information with a user-specified group that includes substantially similar recorded subpopulations as the confirmed cases in the first category;
a sixth step of categorizing the confirmed cases in the third step into a second category of the at least two categories where the second category includes confirmed cases with associated genetic information;
a seventh step to identify and correlate the second category that includes the confirmed cases with associated genetic information with genetic information generated in a bioinformatics process;
an eighth step of executing a machine learning algorithm of a bioinformatics process to assign the correlated confirmed cases in the seventh step to a group of isolates; and
a ninth step of compiling the cases with assigned groups from the second step, the fifth step, and the eighth step into a combined dataset including all isolates assigned to identified groups and associated patient outcome data.

### Claim 15 (depends on 14)

The method of claim 14 wherein the validation process further includes a second set of steps adapted to assess whether the user specified groups reflect risk differences by externally validating the user-specified groups using identified metrics and to optionally revise the number of user-specified groups, the second set of steps comprising the steps of:
compiling the assigned user-specified groups in the combined dataset into a revised database of associated outcomes of the at least one human illness with at least two assigned user-specified groups;
accessing and applying a per-user-specified group calculation of patient outcome metric(s) to the data in the revised dataset;
accessing and applying at least one statistical analysis to the data in the revised dataset;
determining the differences in patient outcome metrics between groups to indicate whether or not a valid and/or desired severity discrepancy according to the at least one applied statistics analysis exists;
compiling validated groups when the differences in patient outcome metrics between groups are within a desired severity discrepancy;
revising the number or composition of the user-specified groups if the at least two user-defined groups outputted from a bioinformatics process are different than the determined differences in epidemiological patient outcome metrics between groups whereby one or more high-risk strains or variants of the desired pathogen are identified; and
if the number or construction of the user-specified groups are revised when the epidemiological patient outcome metrics are different than the user-specified groups identified by the bioinformatics process, one or more iterations of the bioinformatics process and the validation process are executed until the outcome of the validation process is satisfactory.

### Claim 16 (depends on 2)

The method of claim 2 wherein the prioritization process comprises the steps of:
accessing a matrix or dataset of a final set of isolates with each isolate of the final set of isolates having a set of genetic virulence elements and being allocated to one user-specified groups of interest generated by the bioinformatics process and/or a compiled validated set of confirmed cases with assigned groups stored in a subset of a dataset generated via execution of a first set of instructions of the validation process;
accessing and applying user-defined criteria to calculate statistics for each subpopulation to identify a subgroup of pathogens within a group of interest, the user-defined criteria including at least one user-defined statistical threshold;
classifying each subpopulation as “yes” or “no” based upon whether a subpopulation meets a user-defined statistical threshold;
classifying all subpopulations that meet the statistical threshold as a priority subgroup.

### Claim 17 (depends on 9)

The method of claim 9 wherein the prioritization process comprises the steps of:
accessing a matrix or dataset of a final set of isolates with each isolate of the final set of isolates having a set of genetic virulence elements and being allocated to one user-specified groups of interest generated by the bioinformatics process and/or a compiled validated set of confirmed cases with assigned groups stored in a dataset generated via execution of a first set of instructions of the validation process;
accessing and applying user-defined criteria to calculate statistics for each subpopulation to identify a subgroup of pathogens within a group of interest;
classifying each subpopulation as “yes” or “no” based upon whether a subpopulation meets a user-defined statistical threshold;
classifying all subpopulations that meet the statistical threshold as a priority subgroup.

### Claim 18 (depends on 17)

The method of claim 17 wherein the user defined criteria include the frequency in the source of the pathogen; the incidence of human cases; the burden of human cases as calculated using disability adjusted life years; the abundance of genetic elements of interest (e.g., virulence genes); incidence of human cases, relative to presence of subpopulation in the source of pathogen; the abundance of genetic elements of interest, relative to presence of subpopulation in the source of the pathogens.

### Claim 19 (depends on 13)

The method of claim 13 wherein the optimization process includes the steps of:
accessing an input matrix comprising either the output of the biometrics process or the output of the prioritization process;
accessing and applying either an unsupervised machine learning application whereby a group of refined subpopulations of interest are created based upon the input matrix and metrics are generated by which the genetic elements in the input matrix can be compared and ranked or accessing and applying a supervised machine learning algorithm trained on phenotype-differentiating genetic elements to target phenotypic groups and to calculate metrics by which the genetic elements in the input matrix can be compared and ranked;
accessing the refined subpopulations of interest or the target phenotypic groups and compiling a dataset of a ranked and potentially reduced number of genetic elements of the desired pathogen of interest;
optionally, accessing and executing an application or program to determine a number of genetic elements to be used; and
applying an algorithm to find the optimal combination of genetic elements of interest in the dataset of a ranked and potentially reduced number of genetic elements of the desired pathogen of interest or in the optional dataset containing a number of genetic elements to be used.

### Claim 20 (depends on 19)

The method of claim 19 wherein the unsupervised machine learning application comprises an unsupervised random forest analysis.

### Claim 21 (depends on 19)

The method of claim 19 wherein the number of genetic elements to be used may be determined by a user of the system.

### Claim 22 (depends on 19)

The method of claim 19 wherein the application or program optionally accessed and executed to determine a number of genetic elements to be used comprises a receiver operating characteristic (ROC) curve, from which an area under the receiver operating characteristic curve (AUC) may be used to determine a misclassification metric.

### Claim 23 (depends on 22)

The method of claim 22 wherein the misclassification metric may be used to determine the appropriate number of genetic elements to be used.

### Claim 24

A system adapted to execute a method for identifying and grouping pathogenic isolates of hazardous pathogens by genomic features of higher virulence and associated public health risk and identifying corresponding pathogenic biomarkers in isolates thereof, the system comprising:
at least one data center or server adapted to store data that may be used in the method for identifying and grouping pathogens by their public health risk;
at least one processor that is operatively connected to and in logic communication with the data center; and
at least one non-transitory computer readable medium operably connected to and in logic communication with the processor and configured to store and execute one or more diagnostic protocol(s) or analytical methodologies.

### Claim 25 (depends on 24)

The system of claim 24 wherein the computer-readable medium is adapted to store instructions that, when executed by a computer, cause it to perform a method for identifying and grouping pathogenic isolates of hazardous pathogens by genomic features of higher virulence and associated public health risk and identifying corresponding pathogenic biomarkers in isolates thereof.
