Methods for characterizing infections and methods for developing tests for the same
Inventors
Tsalik, Ephraim L. • Giraldo, Ricardo Henao • SCHNELLER, Benjamin Charles • CRENSHAW, Donna G. • Woods, Christopher
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12718905-B2
Publication Date
2026-08-25
Expiration Date
Abstract
The present disclosure provides methods of developing a test to determine if an infection is present or absent, methods for developing a test to determine the etiology of an infection, methods of using tests to determine presence of an infection, such as infections related to sepsis, versus a non-infectious disease/disorder, such as systemic inflammatory response syndrome (SIRS), and methods of using tests to determine the etiology of an infection. Also provided are systems and computer-implemented methods of developing and using tests as set forth herein.
Core Innovation
The disclosure describes detecting the presence or absence of sepsis in a biological sample by measuring gene expression levels and analyzing them with a classifier. The classifier evaluates expression levels using gene signatures selected to differentiate among healthy, non-infectious illness, SIRS, bacterial infection, fungal infection, and viral infection phenotypes. The signature encompasses a 64-gene set and subsets, utilizing normalized expression values in a probit-link form to distinguish sepsis from noninfectious illness.
The disclosure further describes developing a platform-based test by selecting differentially expressed genes across the multiclass phenotypes to generate an infection classifier, including a test for sepsis. Reported performance includes cross-validation ROC AUCs of 0.902–0.972 for sepsis comparisons and 0.925–0.926 for top 64-gene classifiers, alongside compact models and a qRT-PCR implementation. Additionally, genomic, proteomic, or metabolomic measurements identify subtypes, monitor disease and treatment response over time, and generate clinical reports.
Claims Coverage
The two independent claims cover two inventive features: detecting sepsis using a gene-expression classifier and developing a platform-based infection test using selected genes and a classifier.
Detecting sepsis with a gene-expression classifier
One independent claim measures gene expression levels in a biological sample and analyzes them with a classifier to determine the presence or absence of sepsis, or noninfectious illness or inflammation. The classifier uses normalized gene expression values and a probit link function, evaluating scores relative to a cutoff, threshold, range, or band based on signatures differentiating healthy, non-infectious illness, SIRS, bacterial infection, fungal infection, and viral infection.
Developing a platform-based infection test
Another independent claim analyzes gene expression levels from biological sources on a platform, selects genes with differential expression among healthy, non-infectious illness, SIRS, bacterial infection, fungal infection, and viral infection phenotypes, and uses the expression levels of the selected genes to generate a probit-form classifier usable on the platform to determine the presence or absence of infection, including sepsis.
Together, the independent claims cover applying a probit-form gene-expression classifier to detect sepsis in a biological sample and developing a platform-based test using genes selected across multiple infection and inflammatory phenotypes.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Detecting the presence or absence of sepsis in a biological sample and distinguishing sepsis from the absence of sepsis or noninfectious illness or inflammation, evaluated across samples including blood, sputum, nasopharyngeal swabs or washes, synovial fluid, cerebrospinal fluid, bronchoalveolar lavage, and endotracheal aspirates.
Developing and using a platform-based test to determine the presence or absence of infection, including sepsis, bacterial infection, fungal infection, viral infection, or combinations thereof, in a subject.
Identifying disease subtypes, monitoring disease and treatment response over time, and generating clinical reports using genomic, proteomic, or metabolomic measurements.
Interested in licensing this patent?