Methods for diagnosing and treating acute respiratory infections

Abstract

To provide a method for determining the etiology of an acute respiratory infection in a subject.SOLUTION: A method for determining the etiology in a subject having or suspected of having a viral or bacterial infection comprises: (a) preparing a biological sample obtained from the subject; (b) measuring, on a platform, gene expression levels of multiple genes in the biological sample; and (c) inputting the gene expression values or derivatives thereof into one or more classifiers for identifying the viral or bacterial infection.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] Related Applications This application is a continuation of U.S. Provisional Patent Application No. 62 / 187,683, filed July 1, 2015. No. 62 / 257,406, filed November 19, 2015. No. 6,399,433, the disclosures of which are incorporated herein by reference. This shall be accomplished.

[0002] Federal Government Financial Support Description This invention was funded by the National Institutes of Health (NIH ) under federal grant numbers U01AI066569 and P20RR016480; and HHSN266200400064C, and Defense Advance Awarded by the 2016 DARPA Research Projects Agency (DARPA) Federal grant numbers N66001-07-C-2024 and N66001-09-C-2 This invention was made with government support under US Pat. No. 6,332,322. The U.S. Government has certain rights in this invention. [Background technology]

[0003] Acute respiratory infections are common in emergency care settings and are responsible for significant mortality, morbidity and mortality worldwide. Respiratory tract infections or acute respiratory infections (ARIs) are associated with 2 In 2011, it caused 3.2 million deaths worldwide, more than any other cause and 164 million disability-adjusted life years lost (World Health Organization rganization., 2013a, 2013b). In 2012, Lower respiratory tract infections are the fourth leading cause of death among low- and middle-income Lower respiratory tract infections are the leading cause of death in the country (WHO factsheet, These diseases are also prevalent in developed countries. In the United States, in 2010, the Centers for Disease Control and Prevention The Centers for Disease Control (CDC) estimates that pneumonia and influenza alone account for 10% of deaths in the U.S. population. The study determined that the disease caused 15.1 deaths per 100,000 people. Children aged 15 and over are particularly vulnerable to poor outcomes from ARIs. For example, in 2010, Pneumonia accounts for 18.3% of all deaths worldwide in children under 5 years of age (i.e., almost 400,000 deaths).

[0004] Pneumonia and other lower respiratory tract infections are primarily caused by viruses, bacteria, or, less frequently, true Many different pathogens, including bacteria, can be responsible. Among viral pathogens, influenza The number of affected individuals, seasonal severity, and much higher morbidity and mortality rates The constant flow of information about new strains of bacteria (e.g., avian influenza) that cause illness and mortality However, among viral pathogens, Influenza is just one of many that cause serious human diseases. Respiratory syncytial virus (RSV) is a leading cause of hospitalization among children in developed countries during the winter. In 2005, approximately 33% of RSV infections occurred in children under 5 years of age worldwide. million new cases were reported, with 3.4 million cases severe enough to require hospitalization. It is estimated that between 66,000 and 199,000 children die each year from HIV infection alone. In the United States alone, there are approximately 10,000 cases of rheumatoid arthritis each year in the population over 65 years of age. deaths have been linked to RSV infection. In addition to known viral pathogens, novel and emerging History has shown that infectious diseases can emerge at any time and spread worldwide within days or weeks. Recent examples include the SARS coronavirus, which spread from 2003 to 2009. When it first emerged in 1994, it had a 10% mortality rate. More recently, it has been known as Middle East Respiratory Syndrome (MERS). S) The coronavirus continues to threaten to explode in the Middle East and the death toll from it is expected to rise Both of these infections cause respiratory symptoms and may initially be It may be indistinguishable from other ARIs.

[0005] Viral infections cause the majority of ARIs, but bacterial etiologies are also prevalent, especially in lower respiratory tract infections. The specific causes of bacterial ARI vary geographically and clinically. Depending on the situation, these include Streptococcus pneumoniae ( eumoniae, Staphylococcus aureus , Haemophilus influenzae, Chlamydia pneumoniae Chlamydia pneumoniae, Mycoplasma pneumoniae asma pneumoniae), Klebsiella pneumoniae iae, Escherichia coli, and Pseudomonas aeruginosa Identification of these pathogens is based on their presence in culture. These methods rely on the proliferation of bacteria, typically take several days, and are sensitive to the detection of infectious agents. It is difficult to obtain sufficient samples for testing: 1 In a study of 669 patients, only 14% of patients had "good" results producing positive cultures. They were able to provide sputum samples (Garcia-Vazquez et al., 20 04) Clinicians are aware of the limitations of these trials and as a result are anxious Therefore, antibacterial treatment is frequently prescribed without any confirmation of a bacterial infection.

[0006] The ability to rapidly diagnose the etiology of ARI can help determine the optimal treatment for an individual patient. epidemiological surveillance to identify and track outbreaks, and to counter the rising tide of antimicrobial resistance. The issue of appropriate use of antimicrobials to prevent infection has far-reaching social importance at multiple levels. Early and appropriate antimicrobial treatment is a critical global problem. It is well established that antimicrobial therapy improves outcomes in patients with Up to 73% of outpatients with acute respiratory illnesses are prescribed antibiotics, This accounts for approximately 40% of all antibiotics prescribed to adults in this setting. However, it is estimated that only a small minority of these patients require antibacterial treatment (Ca ntrell et al. 2003, Clin. Ther. Jan;24(1 ):170-82). A similar trend is observed in emergency departments. Presence of viral pathogens Microbiological confirmation of infection does not exclude the possibility of concurrent bacterial infection. As a result, antibacterial drugs are often prescribed "just in case." Empiricism that leads to ambiguity has contributed to the upsurge in antimicrobial resistance (Gould, 2009; Kim, & Gallis, 1989), which, as such, is associated with higher mortality, length of hospital stay, and health (Cosgrove 2006, Clin. Infect. D is., Jan 15;42 Suppl 2:S82-9). In addition, the lack of antibiotics Appropriate use can help prevent drug-related adverse effects and other complications, e.g., Clostridium Can lead to Clostridium difficile-associated diarrhea (Zaas et al., 2014).

[0007] Acute respiratory infections are common to many different diseases, including those that are not caused by infection. It is often characterized by persistent, nonspecific symptoms (e.g., fever or cough). Existing diagnostic methods for RI are deficient in many ways. Conventional microbiological tests are poorly due to poor sensitivity and specificity, slow turnaround time, or due to the complexity of the test. (Zaas et al. 2014, Trends Mol Med 20(10):579-88). Current tests that detect certain viral pathogens, e.g. One limitation of multiplex PCR-based assays is that they cannot detect emergent or pandemic virus strains. Influenza pandemics occur when a new virus spreads to a population for which the population has no natural resistance. Influenza pandemics occur when a new virus spreads. For example, the Spanish Flu epidemic of 1918-1919 killed about 20% to 40% of the world's population. % of the population and killed about 50 million people; in 1957-1958, the Asian Flu million people died; 1968-1969 Hong Kong Flu killed about 1 million people; Between 2009 and 2010, approximately 43 million to 89 million people died from swine flu. The Centers for Disease Control and Prevention estimates that 8,870 to 18,300 people died as a result of the disease. The emergence of these new strains is estimated by This poses a challenge to existing diagnostic methods that are not designed to detect Certification may take several days and may be available at specialized testing centers such as state health departments or the CDC. This was especially evident during the 2009 influenza pandemic, when (Kumar & Henrickson 2012, Clin Microb iol Rev 25(2):344-61). Ebola virus disease in West Africa The current pandemic poses a similar challenge. As epidemics are inevitable, it is likely that we will continue to face this problem.

[0008] Further limitations of diagnostic methods using specific viral or bacterial testing paradigms Even if a pathogenic microorganism can be detected, this does not mean that the patient's symptoms are due to the presence of the detected microorganism. This does not prove that the disease is caused by a pathogen. It may be present as part of an individual's normal flora, known as It can be detected by contamination of a sample (e.g., nasal swab or wash). Recently approved multiplex PCR assays, including one that detects Bacillus subtilis, offer high sensitivity; These tests do not distinguish between asymptomatic carriage of the virus and true infection. There is a high rate of asymptomatic viral shedding, especially in children (Jansen et al. al. 2011, J Clin Microbiol 49(7):2631-26 36) Similarly, even if a single pathogen is detected, the disease may not be transmitted by any available or viable treatment. It may be due to a second pathogen for which no tests were available.

[0009] Report: Host gene expression profiles distinguish viral ARI from healthy controls As described (Huang et al. 2011 PLoS Genetics 7(8): e1002234; Mejias et al., 2013; ch et al. 2005 Genes and Immunity 6:588- 595; Woods et al., 2013; AK Zaas et a l., 2013; AK Zaas et al., 2009). However, Most of these are viral infections or bacterial infections. Detection of viral from bacterial ARIs, a more clinically meaningful distinction (Hu, Yu, Crosby, & Storch, 2013; P Arnell et al., 2012; Ramilo et al., 2007 ).

[0010] Thus, current diagnostic methods are unable to distinguish between bacterial and viral infections, as well as non-infectious causes. in the ability to distinguish between symptoms resulting from infections or to identify bacterial and viral co-infections There are limits. Summary of the Invention

[0011] The present disclosure, in part, overcomes many of the limitations of current methods for determining the etiology of respiratory symptoms. The present invention provides a molecular diagnostic test for detecting co-expressed genes or signatures. The host response to infectious agents is detected by measuring and analyzing these genes. The gene expression signatures are identified in humans or animals that exhibit symptoms consistent with acute respiratory infection. or risk of developing an acute respiratory infection (e.g., during an epidemic or endemic disease outbreak). It can be measured in blood samples from humans or animals who are at risk (e.g., presymptomatic). Measurement of the host response as taught herein can be used to assess the efficacy and safety of bacterial ARI, viral ARI, and / or HIV-infected individuals. Distinguish between infectious and non-infectious etiologies and also examine ARI resulting from bacterial and viral coinfections. It can be released.

[0012] This multi-component test operates with unprecedented precision and clinical applicability, making it the go-to test for healthcare providers. The host (subject or patient) response is used to characterize the infectious agent down to the pathogen class level. in individual patients with symptoms that are not specific to the In some embodiments, the result is that the infectious cause of the condition is excluded. , does not detect respiratory viruses or bacterial species (i.e., does not distinguish between viruses and bacteria) (It distinguishes between specific genera or species of viruses or bacteria, but does not distinguish between specific genera or species of viruses or bacteria.) Contains probes or reagents for pathogens and therefore detects only those specific pathogens This provides advantages over current tests which are limited to providing

[0013] One aspect of the present disclosure is a method for determining whether an acute respiratory symptom in a subject is bacterial, viral, or viral in origin. 1. A method for determining whether a biological sample is of infectious or non-infectious origin, comprising: (a) obtaining a biological sample from a subject; Step (b) obtaining a gene expression profile of interest from the biological sample, referred to as a signature. (c) determining by evaluating the expression levels of a predefined set of genes; ) the technology (i.e., platform) used to make such measurements; (d) normalizing the gene expression levels according to the A bacterial classifier with predefined weighting values ​​(coefficients) for each of the genes in the inputting said normalized value into a disease classifier, a disease classifier, and / or a non-infectious disease classifier; (e) determining whether the output of the classifier is a predictor of infection by ... and (f) using the output to determine whether the patient who provided the sample Having an infection of bacterial or viral origin or a non-infectious disease or condition determining whether the combination of The present invention provides a method consisting essentially of

[0014] Another aspect of the present disclosure is a method for determining whether an acute respiratory infection (ARI) in a subject is caused by a bacterial or viral infection. 1. A method for determining whether a biological sample is of infectious or non-infectious origin, comprising: (a) removing a biological sample from a subject; (b) obtaining a subject's gene expression profile from the biological sample; (c) determining the expression levels of said defined set of genes; The technology (i.e., platform) used to perform the measurements (d) normalizing the gene expression levels to generate normalized values; The normalized values ​​are input into a classifier that has predefined weighting values ​​for each of the genes. (e) computing the output of the classifier based on a defined threshold, cut-off value, or likelihood of infection; (f) comparing said sample to a range of values ​​indicative of the presence or absence of bacteria if said sample is negative for bacteria; repeat step (d) using only the viral and non-infectious classifiers. and (g) a screen for classifying said sample as having a viral etiology or a non-infectious disease. The present invention provides a method comprising, consisting of, or consisting essentially of the steps of:

[0015] Another aspect of the present disclosure is a method for determining whether an acute respiratory infection (ARI) in a subject is caused by a bacterial or viral infection. 1. A method for determining whether a biological sample is of infectious or non-infectious origin, comprising: (a) removing a biological sample from a subject; (b) obtaining a subject's gene expression profile from the biological sample; (c) determining the expression levels of said defined set of genes; The technology (i.e., platform) used to perform the measurements (d) normalizing the gene expression levels to generate normalized values; The normalized values ​​are input into a classifier that has predefined weighting values ​​for each of the genes. (e) computing the output of the classifier based on a defined threshold, cut-off value, or likelihood of infection; (f) comparing said sample to a range of values ​​indicative of a positive result for the virus; If so, repeat step (d) using only the bacterial and non-infectious classifiers. (g) classifying said sample as having a bacterial etiology or a non-infectious disease. The present invention provides a method comprising, consisting of, or consisting essentially of the steps of:

[0016] Another aspect of the present disclosure is a method for determining whether an acute respiratory infection (ARI) in a subject is caused by a bacterial or viral infection. 1. A method for determining whether a biological sample is of infectious or non-infectious origin, comprising: (a) removing a biological sample from a subject; (b) obtaining a subject's gene expression profile from the biological sample; (c) determining the expression levels of said defined set of genes; The technology (i.e., platform) used to perform the measurements (d) normalizing the gene expression levels to generate normalized values; The normalized values ​​are input into a classifier that has predefined weighting values ​​for each of the genes. (e) computing the output of the classifier based on a defined threshold, cut-off value, or likelihood of infection; (f) comparing said sample to a range of values ​​indicative of a negative result for a non-communicable disease; If so, repeat step (d) using only the viral and bacterial classifiers. and (g) classifying the sample as being of viral or bacterial etiology. A method is provided that comprises, consists of, or consists essentially of the steps.

[0017] Yet another aspect of the present disclosure is a method for treating acute respiratory infection (ARI) of unknown etiology in a subject. (a) obtaining a biological sample from a subject; (b) administering to the subject a biological sample; and Gene expression profiles of interest from a target sample are collected using a predefined set of genes (e.g., The expression levels of two, three, or more signatures of (c) determining the technology (i.e., the normalizing the gene expression levels for the given platform to generate normalized values; d) A feature vector having predefined weighting values ​​for each of the genes in each signature. Inputting the normalized values ​​into the bacteria classifier, the virus classifier, and the non-infectious disease classifier. (e) determining an output of the classifier based on a predefined threshold, cutoff value, or likelihood of infection; (f) comparing said sample to a range of values ​​for bacterial etiology, viral etiology, or (g) classifying the malaria disease identified by step (e) as a non-communicable disease; administering to said subject an appropriate treatment regimen as described in or a method consisting essentially of the steps. (g) includes administering antibacterial therapy if the etiology of the ARI is determined to be bacterial. In another embodiment, step (g) comprises determining whether the etiology of the ARI is viral. In some cases, this includes administering antiviral therapy.

[0018] Another embodiment is an acute respiratory infection selected from bacterial, viral, and / or non-infectious. Response to a vaccine or drug in subjects suffering from or at risk for a disease 20. A method for monitoring a subject's ability to differentiate from a host, comprising determining a host response in said subject, said determining The steps are carried out according to the methods taught herein. In some embodiments, the drug is an antibacterial or antiviral drug.

[0019] In some embodiments of the above aspects, the method includes generating a score indicative of the probability of ARI etiology. The method further includes generating a report that assigns the items to the targets.

[0020] Acute respiratory infection in subjects selected from bacterial, viral, and / or non-infectious A system for determining the etiology of a disease, comprising: at least one processor; a sample input circuit configured to receive a biological sample; and configured to determine gene expression levels in said biological sample. analysis circuitry; input / output circuitry coupled to said at least one processor; a processor coupled to said processor and configured to store data, parameters, and / or classifiers; a memory circuit coupled to the processor and configured to When executed by at least one processor, the method causes the at least one processor to perform an operation. A memory containing computer readable program code embodied in the memory, (including combinations thereof), wherein said operation comprises: The measurement of gene expression levels of a predefined set of genes (i.e., a signature) in a and controlling / performing the sample analysis circuitry; normalizing the gene expression levels to normalize the gene expression levels. generating a defined gene expression value for each of the genes in the defined gene set; Bacterial acute respiratory infection (ARI) classifier with predefined weighting values ​​(i.e., coefficients) retrieving a viral ARI classifier and a non-infectious illness classifier from said storage circuit. Bacterial acute respiratory infection (ARI) classifiers, viral ARI classifiers, and non-infectious said normalizing gene expression to one or more acute respiratory disease classifiers selected from disease classifiers. and inputting a value for each of the classifiers; calculating a probability of etiology for one or more of the infectious diseases; and Whether the acute respiratory illness in the country is of bacterial, viral, or non-infectious origin, or any of these and controlling the output of a decision among several combinations via the input / output circuit. A stem is further provided.

[0021] In some embodiments, the system provides quantitative or semi-quantitative detection of gene expression. The computer readable code converts the ARI etiology into a cumulative score or probability.

[0022] In some embodiments, the system includes an array platform, a thermal cycler, Larger platforms (e.g., multiplexing and / or real-time PCR platforms) ), hybridization and multi-signal coded (e.g., fluorescent) detectors Platform, Nucleic Acid Mass Spectroscopy Platform, Nucleic Acid Sequencing Platform or a combination thereof.

[0023] In some embodiments of the above aspects, the defined gene set comprises at least three Contains gene signatures of

[0024] In some embodiments of the above aspects, the biological sample is peripheral blood, sputum, nasopharyngeal swab, or the like. , nasopharyngeal wash, bronchoalveolar lavage fluid, endotracheal aspirate, and combinations thereof. The sample includes a sample selected from the following:

[0025] In some embodiments of the above aspects, the bacterial classifier is selected from Table 1, Table 2, Table 9, Table 10, and / or a gene listed as part of the bacterial classifier in Table 12 (e.g., (can be measured using oligonucleotide probes that are homologous to genes or gene transcripts) 5, 10, 20, 30, or 50 to 80, 100, 150, or comprises up to 200 expression levels. In some embodiments, the viral classifier comprises As part of the virus classifier in Table 1, Table 2, Table 9, Table 10, and / or Table 12 The listed genes (e.g., oligonucleotides homologous to the genes or gene transcripts) 5, 10, 20, 30, or 50 out of 8 of the 5, 10, 20, 30, or 50 that can be measured with the In some embodiments, the expression level may be 0, 100, 150, or up to 200. In this embodiment, the non-communicable disease classifier is selected from Table 1, Table 2, Table 9, Table 10, and / or Table 12. Genes listed as part of the non-infectious disease classifier in Five of the ten genes were identified using oligonucleotide probes homologous to gene transcripts. pcs, 20 pcs, 30 pcs, or 50 pcs to 80 pcs, 100 pcs, 150 pcs, or 200 pcs Including expression levels at

[0026] (a) means for extracting mRNA from a biological sample; (b) the methods described in Table 1, Table 2, Table 9, Table 10, and / or 8 out of 5, 10, 20, 30, or 50 from Table 12 Regions of homology to transcripts from 0, 100, 150, or up to 200 genes For producing one or more arrays of synthetic oligonucleotides having and (c) comprising, consisting of, or consisting essentially of instructions for use. Also provided is a kit for determining the etiology of an acute respiratory infection (ARI) in a subject. will be done.

[0027] Another aspect of the present disclosure is the use of a kit to evaluate acute respiratory infection (ARI) classifiers. (a) a method of using five compounds from Table 1, Table 2, Table 9, Table 10, and / or Table 12, 10, 20, 30, or 50 to 80, 100, 150, or 2 A single synthetic oligonucleotide sequence consisting of multiple synthetic oligonucleotides with regions of homology to up to 100 genes. (b) preparing an array or arrays of oligonucleotides having regions homologous to normalization genes; (c) adding nucleotides to the array; (d) obtaining a biological sample from a subject having a genomic DNA sequence; (e) isolating RNA from said sample and (e) generating a transcriptome (e.g., proportional to the level of gene expression); and measuring the transcriptome on the array (e.g., by measuring fluorescence or current). (f) measuring the transcriptome measurements against normalization genes; Normalizing the measurements and electronically transferring the normalized measurements to a computer for running the classifier. (g) generating a report, and optionally, (h) making an appropriate decision based on said results. The present invention provides a method comprising, consisting of, or consisting essentially of administering a treatment do.

[0028] In some embodiments, the method further comprises: The method further includes externally validating the ARI classifier against a known dataset. In one embodiment, the datasets are GSE6269, GSE42026, GSE40 396, GSE20346, GSE42834, and combinations thereof. It is selected.

[0029] Yet another aspect of the present disclosure provides all that is disclosed and exemplified herein.

[0030] The present invention relates to a method of treatment for acute respiratory infection (ARI) in a subject of unknown etiology. Use of the ARI classifiers as taught herein is also provided.

[0031] The foregoing aspects and other features of the present disclosure are illustrated herein in conjunction with the accompanying drawings, in which: This is explained in the following description. [Brief description of the drawings]

[0032] [Figure 1] 1 is a schematic diagram showing a method (training 10) of obtaining classifiers according to some embodiments of the present disclosure, where each classifier is composed of a weighted sum of all or a subset of normalized gene expression levels. This weighted sum defines the probability that allows a decision (classification), especially when compared with a threshold or confidence interval. The exact combination of genes, their weights, and thresholds for each classifier obtained by training are specific to a particular platform. The classifier (or more precisely, its components, i.e., weights and thresholds or confidence intervals (values)) becomes a database. Weights with non-zero values ​​determine the subset of genes used by the classifier. Repeat to obtain all three classifiers (bacterial ARI, viral ARI, and non-infectious ARI) within the identified platform that match gene expression values. [Diagram 2] FIG. 1 illustrates an example of creating and / or using a classifier according to some embodiments of the present disclosure. [Diagram 3] FIG. 1 is a schematic diagram showing a method for classifying 20 the etiology of an acute respiratory symptom suffered by a subject using a classifier according to some embodiments of the present disclosure. [Figure 4] FIG. 1 is a schematic diagram showing a decision pattern for using secondary classification to determine the etiology of an ARI in a subject according to some embodiments of the present disclosure. [Diagram 5] FIG. 1 is a diagram of an example of the training method presented in Example 1. A cohort of patients with bacterial ARI, viral ARI, or non-infectious disease was used to develop a classifier for each condition. This integrated ARI classifier was validated using leave-one-out cross-validation and compared with three published classifiers of bacterial infection vs. viral infection. The integrated ARI classifier was also externally validated in six publicly available datasets. In one experiment, healthy volunteers were included in the training set to determine their suitability as "non-infected" controls. All subsequent experiments were performed without this healthy subject cohort. [Figure 6]Graph showing the results of leave-one-out cross-validation of three classifiers (bacterial ARI, viral ARI, and non-infectious illness) according to the example training method presented in Example 1. Each patient is assigned a probability of having bacterial ARI (triangle), viral ARI (circle), and non-infectious illness (square). Patients who were clinically determined to have bacterial ARI, viral ARI, or non-infectious illness are presented in the top, middle, and bottom panels, respectively. The overall classification accuracy was 87%. [Figure 7] 1 is a graph showing the evaluation of healthy adults as non-infected controls rather than diseased but uninfected controls. This figure demonstrates the unexpected superiority of using diseased but uninfected subjects as controls. [Figure 8] FIG. 1 shows positive and negative predictive values ​​for A) bacterial and B) viral ARI classification as a function of prevalence. [Figure 9] Venn diagram showing overlap in the bacterial, viral, and non-infectious illness classifiers. There are 71 genes in the bacterial ARI classifier, 33 genes in the viral ARI classifier, and 26 genes in the non-infectious illness classifier. One gene overlaps between the bacterial and viral ARI classifiers. Five genes overlap between the bacterial ARI classifier and non-infectious illness classifiers. Four genes overlap between the viral ARI classifier and non-infectious illness classifiers. [Figure 10]Figure 1 shows classifier performance in patients with co-infection with identification of bacterial and viral pathogens. The bacterial and viral ARI classifiers were trained on subjects (GSE60244) with bacterial (n=22) or viral (n=71) infections. This same dataset also included 25 subjects with bacterial / viral co-infection. As shown in the figure, the bacterial and viral classifier predictions were normalized to the same scale. Each subject receives two probabilities: a probability of bacterial ARI host response and a probability of viral ARI host response. Probability scores of 0.5 or higher were considered positive. Subjects 1-6 have a bacterial host response. Subjects 7-9 have both bacterial and viral host responses that may indicate true co-infection. Subjects 10-23 have a viral host response. Subjects 24-25 have neither a bacterial nor a viral host response. [Figure 11]11 is a block diagram of a classification system and / or computer program product that may be used in a platform. The classification system and / or computer program product 1100 may include a processor subsystem 1140 including one or more central processing units (CPUs) on which one or more operating systems and / or one or more applications run. Although one processor 1140 is shown, it will be understood that there may be multiple processors 1140 that may be either electrically interconnected or separate. The processor 1140 is configured to execute computer program code from a memory device such as memory 1150 to perform at least a portion of the operations and methods described herein. The storage circuitry 1170 may store a database that provides access to data / parameters / classifiers used by the classification system 1100, such as signatures, weights, thresholds, etc. The input / output circuitry 1160 may include a display and / or user input devices such as a keyboard, touch screen, and / or pointing device. Devices associated with the input / output circuitry 1160 may be used to provide information to the processor 1140 by a user of the classification system 1100. Devices associated with the input / output circuit 1160 may include a network or communication controller, input devices (keyboard, mouse, touch screen, etc.), and output devices (printer or display). An optional update circuit 1180 may be included as an interface for providing updates to the classification system 1100, such as updates to the code executed by the processor 1140 stored in the memory 1150 and / or the storage circuit 1170. Updates provided via the update circuit 1180 may also include updates to portions of the storage circuit 1170 associated with databases and / or other data storage formats that maintain information about the classification system 1100, such as signatures, weights, thresholds, etc. The sample input circuit 1110 provides an interface for the classification system 1100 to receive a biological sample to be analyzed.The sample processing circuit 1120 may further process the biological sample within the sorting system 1100 to prepare the biological sample for automated analysis. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0033] For the purposes of promoting an understanding of the principles of the present disclosure, reference will now be made to preferred embodiments, Specific language is used to describe it. However, it is understood that the limitations of the scope of the disclosure are not to be construed as limiting the scope of the invention. No such limitation is intended by the present disclosure as exemplified herein. It is understood that variations and further modifications are contemplated as would normally occur to one of ordinary skill in the art. It will be possible.

[0034] The articles "a" and "an" refer to one or more of the grammatical objects of the article. is used herein to refer to more than one (ie, at least one). By way of example, "an element" means at least one element and includes more than one element. obtain.

[0035] Unless otherwise specified, all technical terms used herein are understood to be within the meaning of the present invention by those of ordinary skill in the art. has the same meaning as understood in

[0036] The present disclosure provides a method for identifying genes, proteins, and proteins in blood in response to exposure to pathogens that cause acute respiratory infections. Changes in protein and metabolite expression can identify with a high degree of accuracy the etiology of ARI in subjects. The present invention provides methods that can be used to determine and characterize

[0037] definition As used herein, the term "acute respiratory infection" or "ARI" refers to a respiratory infection that is often Symptoms consistent with infection of the upper or lower respiratory tract with a bacterial or viral pathogen; and / or physical findings (e.g., symptoms such as coughing, wheezing, fever, sore throat, congestion; rapid heart rate, Hypertension, elevated respiratory rate, abnormal white blood cell count, low arterial carbon dioxide pressure (PaCO2), etc. Symptoms are characterized by rapid progression of symptoms over hours to days. ARI refers to an infectious disease or illness that occurs primarily in the upper respiratory tract (URI), lower respiratory tract (LRI), or a combination of the two. ARI is defined as the spread of infection beyond the respiratory tract. Systemic effects may occur due to the presence of steroids or collateral damage induced by the immune response. Examples of the former include Staphylococcus aureus pneumonia that has spread to the bloodstream and endocarditis (sensitivity of the heart valves). may result in secondary sites of infection, including septic arthritis (joint infection), or osteomyelitis (bone infection). An example of the latter is influenza pneumonia resulting in acute respiratory distress syndrome and respiratory failure. include.

[0038] As used herein, the term "signature" refers to a combination of genes that are unique to a particular organism. A set of biological analytes and measurements of said analytes that indicate the presence or absence of a biological condition. These signatures have known states (e.g., confirmed breathing patterns). have multiple infections (including respiratory bacterial infections, respiratory viral infections, or non-communicable diseases) One or more categories or outcomes of interest have been found in elephants ( These measurable substances, also known as biological markers, are used to identify Analytes of interest may include gene expression levels, protein or peptide levels, or metabolite levels. The present invention may be (but is not limited to) a microprocessor. Publication No. 2015 / 0227681; U.S. Patent Publication No. 2016 / 01539 to Eden et al. See also issue 93.

[0039] In some embodiments as disclosed herein, a "signature" is a When the current levels are incorporated into a classifier as taught herein, bacterial ARI Specific combinations of genes that identify conditions such as viral ARI or non-communicable disease See, for example, Tables 1, 2, 9, 10, and 12 below. In an embodiment, the signature is agnostic to respiratory viruses or bacterial species ( That is, a distinction is made between viruses and bacteria, but this is only between specific genera of viruses or bacteria. or species), and / or are insensitive to specific causes of non-communicable disease .

[0040] As used herein, the terms "classifier" and "predictor" are used interchangeably. r) are used interchangeably and refer to a given observation for the purposes of assigning it to a category. The values ​​of the signatures (e.g., defined Gene expression levels for the selected gene set and the determined A classifier is a mathematical function that uses linear and / or probabilistic The score may be a function of the sum of the signature values ​​weighted by a set of coefficients. Furthermore, if a function of the signature values ​​yields a probability, then ,The classifier is probabilistic, and a value between 0 and 1.0 (or 0% and 100%) represents the Each of these is a set of assumptions that an object or observation belongs to a particular category or will produce a particular result. Quantify the likelihood. Probit and logistic regression analyses produce probabilities. To do this, we use the probit function and the logistic link function, respectively. This is an example of a linear classifier.

[0041] A classifier as taught herein is obtained by a procedure known as "training." The procedure can be performed based on known category membership (e.g., bacterial ARI, Data sets containing observations on infectious, viral, and / or non-communicable diseases See Figure 1. Specifically, training involves the use of a predefined signature in addition to the optimal signature. The optimal coefficient (i.e., weight) for each component (e.g., gene expression level component) of the ), with the optimal outcome being determined by the highest achievable classification accuracy. .

[0042] A "classification" is a classification of subjects who are suffering from or at risk of suffering from acute respiratory symptoms. or multiple categories or outcomes (e.g., a patient is infected or not infected with a pathogen) Another categorization is whether the patient is infected with a virus and / or a bacteria. See FIG. 3. In some cases, the subject may It may fall into more than one category (eg, in the case of bacterial and viral co-infection). The outcome or category is determined by the score value provided by the classifier. The value may be compared to a cut-off value or threshold, a confidence level or limit. For each category, the probability of belonging to a particular category may be given (e.g., if the classifier reports a probability, ).

[0043] As used herein, the term "indicates" when used in conjunction with gene expression levels. "dicative" means that gene expression levels are comparable to those of alternative biological states (e.g., bacterial up-regulation compared to expression levels in either ARI or viral ARI or controls This means that the protein level is either down-regulated, altered, or changed. The term "indicates" when used in conjunction with means that the protein levels are indicative of a standard protein level or is higher or lower, increased or decreased, compared to levels in alternative biological states. It means slightly altered or changed.

[0044] The terms "subject" and "patient" are used interchangeably and refer to a subject being examined, studied, or treated. It is understood that the present disclosure is not limited to any particular type of subject. It is not intended that humans be the preferred subjects in some embodiments of the present invention, but other In this embodiment, non-human animals are preferred subjects, including mice, monkeys, Ferrets, cattle, sheep, goats, pigs, chickens, turkeys, dogs, cats, and cows. In certain embodiments, the mammals include, but are not limited to, mammals, horses, and reptiles. The subject is suffering from an ARI or exhibiting ARI-like symptoms.

[0045] As used herein, "platform" or "technology" refers to any technology that is part of the present disclosure. According to the present invention, a device that can be used to measure a signature, e.g., a gene expression level ( For example, instruments and associated components, computers, one or more refers to a computer readable medium that contains a plurality of databases, reagents, etc. Examples of systems include array platforms, thermal cycler platforms (e.g., Multiplexing and / or real-time PCR platforms), nucleic acid sequencing platforms Platforms, hybridization and multi-signal coding (e.g., fluorescence) Detector platforms, such as nucleic acid mass spectrometry platforms and magnetic resonance platforms Examples of suitable cellular components include, but are not limited to, cellular components, cellular components, and combinations thereof.

[0046] In some embodiments, the platform semi-quantitatively measures gene expression levels. The present invention is designed to determine the expression of a specific gene, i.e., not to measure its individual or absolute expression. , expression levels are used as estimates and / or relative to each other or to the identified markers ( For example, the expression of another "standard" or "reference" gene.

[0047] In some embodiments, the semi-quantitative measurement includes an estimate of the gene expression in the signature or To provide relative expression levels, signals representing identified mRNAs were detected. The PCR cycles are performed at 100 s, and the number of PCR cycles required until detection is used. "Time PCR" is included.

[0048] Real-time PCR platforms include, for example, TaqMan® Low Density The array (TLDA) includes a real-time PCR array in which samples are On an array card having a collection of wells to be subjected to multiplex reverse transcription, followed by real-time The results are analyzed by real-time PCR. Kodani et al. 2011, J. Clin. Microbiol. 49(6):2175-2182. Platforms also include, for example, Biocartis Idylla™ sample preparation It includes a technology from start to finish, in which cells are lysed and NA / RNA is extracted, real-time PCR is performed, and the results are detected.

[0049] Magnetic resonance platforms include, for example, the T2 Biosystems® T The technology includes Transducer 2 Magnetic Resonance (T2MR®) technology, Thus, molecular targets can be identified in biological samples without the need for purification.

[0050] The terms "array", "microarray", and "microarray" are interchangeable; Refers to an arrangement of a collection of nucleotide sequences displayed on a support. Any type of array can be utilized in the methods provided herein. For example, the array can be Solid supports such as glass (solid-phase arrays) or semi-solid supports such as nitrocellulose membranes The array may also be presented on beads, i.e., a bead array. These beads are typically microscopic and made, for example, of polystyrene. The arrays can also be presented on nanoparticles, which can be, for example, in particular gold. However, they can also be made of silver, palladium, or platinum. For example, gold nanoparticle probes Nanosphere Verigene® System Using Nanosphere Technology Magnetic nanoparticles may also be used. Other examples include nuclear magnetic resonance microcoils. Nucleotide sequences can be DNA, RNA, or any permutation thereof. tation) (e.g., nucleotide analogs such as locked nucleic acids (LNA)). In some embodiments, the nucleotide sequence may be derived from a sequence selected from the genome rather than from a genome-wide sequence. Exon / intro sequencing to detect gene expression of spliced ​​or mature RNA species Nucleotide sequences can also be derived from partial sequences, primers, or whole genes. The sequence may be a non-coding sequence, a coding sequence, a published sequence, a known sequence, or a novel sequence. The array may additionally comprise antibodies, It may include other compounds such as peptides, proteins, tissues, cells, chemicals, carbohydrates, etc.

[0051] Array platforms include, for example, the TaqMan® low-density arrays mentioned above. Translocation density array (TLDA), and Affymetrix® microarray platform This includes the form.

[0052] Hybridization and multi-signal coding detector platforms include: For example, NanoString nCounter® technology (e.g. A color code attached to a target-specific probe (corresponding to the gene expression transcript of interest) (detection of hybridization of the chromatinized barcodes) and Luminex (registered xMAP® Technology (microsphere beads are color coded and coated with a target (e.g., gene expression transcript) specific probe for detection. ), and Illumina® BeadArray (microbeads) The target for detection (e.g., a gene) is collected onto a fiber optic bundle or a flat silica slide. The plasmid contains a probe coated with a gene expression transcript specific probe.

[0053] Nucleic acid mass spectrometry platforms include, for example, Ibis Biosciences P and a lex-ID® detector in which DNA mass spectrometry is performed by: It is used to detect amplified DNA using its mass profile.

[0054] Thermal cycler platforms include, for example, the FilmArray® Multiplex PCR system (extracting and purifying nucleic acid from raw samples and performing nested multiplex PCR) (to carry out PCR), and the RainDrop Digital PCR System (to carry out PCR using a microfluidic chip) The present invention relates to a droplet-based PCR platform that uses a 3D printer.

[0055] The term "computer-readable medium" refers to a medium that stores information (e.g., data and instructions) and that Refers to any device or system for supplying data to a computer processor. Examples of data-readable media include DVDs, CDs, hard disk drives, magnetic tapes, and netbooks. Streaming media and applications over networks (e.g., smartphones) Servers for mobile devices (such as those found on smartphones and tablets) In various embodiments, aspects of the invention, including data structures and methods, The processing and data may also be stored on a number of device types. The device types include desktop and laptop computers, This includes, but is not limited to, bullets, smartphones, etc.

[0056] As used herein, the term "biological sample" refers to a sample obtained from a subject in the methods provided herein. "Genetic samples" includes any sample that can be taken from a subject that contains genetic material that can be used in the present invention. For example, the biological sample can include a peripheral blood sample. The term "peripheral blood sample" refers to blood that is collected from the circulatory system. The term refers to blood circulating in a blood vessel or a sample of blood taken from that system of the body. Other samples are These may include samples taken from the respiratory tract, including sputum, nasopharyngeal swabs, and nasopharyngeal washings. Biological samples also include, but are not limited to, samples taken from the lower respiratory tract. may include, but are not limited to, bronchoalveolar lavage fluid and endotracheal aspirate. A biological sample may also include any combination thereof.

[0057] The term "genetic material" refers to the storage of genetic information in the nucleus or mitochondria of the cells of an organism. Examples of genetic material include double-stranded and single-stranded DNA, cDNA, These include, but are not limited to, NA, RNA, and mRNA.

[0058] The term "multiple nucleic acid oligomers" refers to two or more nucleic acid oligomers, which may be DNA or RNA. Refers to Ligomer.

[0059] As used herein, the terms "treat", "treatment", and "treating" refer to , a disease or disorder or one or more symptoms resulting from the administration of one or more treatments Such terms refer to a decrease or improvement in the severity, duration, and / or progression of symptoms. , a decrease in viral or bacterial replication, or other symptoms in a viral or bacterial subject. Treatment also refers to the reduction of spread to an organ or tissue, or to other subjects. These may include therapies for ARIs due to non-infectious diseases, such as anti-inflammatory drugs, anti-inflammatory drugs, and anti-inflammatory drugs.

[0060] The term "effective amount" refers to an amount of a therapeutic agent that is sufficient to exert a physiological effect in a subject. The term "responsiveness" refers to the level of gene expression of a gene in a subject when said subject is infected with a virus. In response to infection with a virus, bacteria, or a non-communicable disease, Subjects without bacterial infection or non-communicable disease, or control subjects The term refers to a change in the gene expression level of the gene in a given condition compared to the gene expression level in a given condition.

[0061] The term "appropriate treatment regimen" refers to the care required to treat a particular disease or disorder. Such a regimen is capable of producing a curative effect in a disease state. In many cases, this requires administering a therapeutic agent to a subject. For example, Therapeutic agents for treatment are antibiotics, including penicillins, cephalosporins, These include fluoroquinolones, tetracyclines, macrolides, and aminoglycosides. Therapeutic methods for treating subjects with viral respiratory infections include, but are not limited to, Therapeutic agents include oseltamivir, RNAi antivirals, inhaled ribavirin, and monoclonal antibodies. These include respigam, zanamivir, and neuraminidase blocking agents. The present invention is not limited to the use of antiviral or antibiotic treatments not yet available. The present invention contemplates the use of the methods of the present invention to determine the presence or absence of a non-infectious Treatment of ARI due to disease includes, but is not limited to, anti-inflammatory drugs. Staminics, decongestants, anticholinergic nasal sprays, leukotriene inhibitors, mast cell inhibitors allergy treatments, including, but not limited to, administration of inhaled contraceptives, steroid nasal sprays, and the like; Leukosteroids, leukotriene modifiers, long-acting beta-agonists, combination inhalers (e.g. , fluticasone-salmeterol; budesonide-formoterol; mometasone-formoterol rol), theophylline, short-acting beta agonists, ipratropium, oral and intravenous and asthma treatments including intravenous corticosteroids, omalizumab, and the like.

[0062] Such regimens involve the administration of therapeutic agents capable of producing a reduction in symptoms associated with a disease state. Examples of such drugs include NSAIDS, Acetaminophen, antihistamines, beta agonists, cough suppressants, or symptoms associated with a disease process These include, but are not limited to, other drugs that reduce the symptoms of

[0063] How to create a classifier (training) The present disclosure provides a method for generating classifiers for use in a method for determining the etiology of acute respiratory illness in a subject. The present invention provides a method for determining the etiology of ARI in a subject with a high degree of accuracy (also referred to as training 10). A gene expression-based classifier has been developed that can be used to identify and characterize It has been issued.

[0064] Thus, as shown in FIG. 1, one embodiment of the present disclosure is a method for treating acute respiratory infections (AR). I) A method for generating a classifier comprising: (i) determining whether a bacterial, viral, or non-infectious acute respiratory syndrome is present; A procedure for obtaining biological samples (e.g., peripheral blood samples) from multiple subjects suffering from a respiratory infection. (ii) optionally, generating RNA (e.g., a transcriptome) from the sample; (iii) isolating total RNA (for preparation of DNA) (105, not shown in FIG. 1); A number of genes (i.e., some or all of the genes expressed in the RNA) (iv) a step of normalizing the gene expression levels; and (v) a bacterial ARI classifier, a viral ARI classifier, and generating a classifier for a disease caused by a disease that is ... Or a method consisting essentially of them.

[0065] In some embodiments, the sample is not purified after collection. In some cases, the sample may be purified to remove extraneous material before or after lysis of the cells. In some embodiments, the sample is subjected to cell lysis and removal of cellular material, isolation of nucleic acid, and / or reduction of abundant transcripts such as globin or ribosomal RNA. It is then refined.

[0066] In some embodiments, measuring the gene expression level comprises measuring the level of the transcript. generating one or more microarrays using the primers; measuring the transcriptome using a genomic DNA sequence; analyzing and correcting for batch differences The method may include steps.

[0067] In some embodiments, the method further comprises: The associated weight (W n ), and upload thresholds to one or more databases. The method further includes a step 140 of downloading the

[0068] An example of creating the classifier is detailed in FIG. 2. As shown in FIG. Biology from cohorts of patients including ARI, viral ARI, or non-communicable illness Targeted samples were collected for each condition (i.e., bacterial acute respiratory infection, viral acute respiratory infection, to develop gene expression-based classifiers for infectious or non-infectious diseases. Specifically, the bacterial ARI classifier is used to classify bacterial ARIs into either viral ARIs or non-infectious diseases. These are obtained to reliably identify individuals with bacterial ARI against either the The viral ARI classifier classifies viral ARIs as bacterial ARIs or non-infectious illnesses (NIs). Non-infectious disease classifiers are used to reliably identify those with RI. The first is to improve the specificity of the bacterial ARI classifier and the viral ARI classifier. The signatures for the classifier, viral ARI classifier, and non-infectious disease classifier are ( For example, by applying a sparse logistic regression model.

[0069] These three classifiers are then combined, if necessary, to find the class with the maximum membership probability. A single, named "ARI classifier" is created by following a one-to-many scheme to specify the The integrated ARI classifier may be integrated into the classifier. See also FIG. 5. The integrated ARI classifier may be integrated into the classifier in some embodiments. and / or may be validated using leave-one-out cross-validation in the same population from which it was derived; Alternatively, in some embodiments, samples from subjects suffering from a disease of known etiology. Validation can be performed using publicly available human gene expression datasets. For example, validation can be performed using Publicly available human gene expression datasets (e.g., GSE6269, GSE420 26, GSE40396, GSE20346, and / or GSE42834) The data set can be carried out in a multicenter, multicenter, and multicenter study, and the data set can be organized into two groups: bacterial ARI, The studies were selected based on the condition that they included all major infectious diseases (viral ARI, or non-communicable disease).

[0070] The classifiers are classified into three groups: disease of known etiology, i.e., bacterial ARI, viral ARI, or non- It can be validated in a standard set of samples from subjects suffering from an infectious disease.

[0071] The methodology for training described herein can be used by those skilled in the art to detect different gene expression ( For example, the present invention can be easily translated into a novel mRNA detection and quantification platform.

[0072] The disclosed methods and assays are based on gene expression, e.g., direct measurement of RNA. , measurement of derived material (e.g., cDNA) and RNA products (e.g., coding sequences). Gene expression can be measured through the measurement of protein or peptide. Any method of screening may be used and is within the scope of this disclosure.

[0073] In some embodiments, the measurement includes detection and quantification (e.g., detection of and quantification of mRNA in a sample. In some embodiments, gene expression levels include one or more The levels are adjusted against multiple standard gene levels ("normalization"). As discussed above, normalization involves the use of a process that gives the amount or relative amount (e.g., of expressed genes). This is done to remove technical variability inherent to the platform.

[0074] In some embodiments, detection and quantification of mRNA begins with reverse transcription and / or Alternatively, an amplification step may be included, for example RT-PCR, such as quantitative RT-PCR. In some embodiments, detection and quantification involves detecting a nucleotide sequence present in or from a biological sample. Direct detection and measurement of RNA molecules can be based on purified, unamplified mRNA molecules. Typically, hybridization with a complementary primer and / or a labeled probe is performed. Such methods include traditional Northern blotting and surface-enhanced Raman spectroscopy. SERS (Single-Element Resonance Imaging) method, which involves the surface of a plasmon-active metal structure carrying gene-specific probes The researchers shine a laser on a sample exposed to light and measure the change in light frequency as it scatters. (including the

[0075] Similarly, detection of RNA derivatives such as cDNA typically involves the use of complementary primers and and / or hybridization with labeled probes. nucleotide probe arrays (e.g., solid-state microarrays and bead arrays) or related Probe-hybridization methods and the relative and absolute determination of specific RNA molecules PCR methods, including real-time, digital, and end-point PCR methods for paired quantification PCR-based amplification and detection may be included.

[0076] Additionally, sequencing-based methods can be used to quantify RNA or RNA-derived material levels. When applied to RNA, sequencing can be used to detect and quantify The sequencing method, called RNAseq, provides qualitative information about the RNA molecules in a sample ( of RNA or its cognate cDNA) and quantitative information (copy For example, Wang et al. 2009 Nat. Rev. See Genet. 10(1):57-63. Another sequence-based method, Serial Analysis of Age Expression (SAGE) is a surrogate for measuring the expression levels of RNA molecules. Using cDNA "tags."

[0077] In addition, the use of a proprietary platform for mRNA detection and quantification is also included herein. These examples can be used to implement the disclosed methods. Molecular Indexing (trademark) developed by ARCH, INC. Incorporated Pixel™ System, NanoString™ T echnologies nCounter Gene Expression System; Illumina mRNA-Seq, Tag-Profiling, BeadArray™ Technology G, VeraCode, PrimeraDx's ICEPlex system, and Affymetrix's QuantiGene 2.0 multiplex assay.

[0078] As an example, RNA from whole blood from a subject is purified using PAXgene™ RNA tubes ( PreAnalytiX, Valencia, Calif.) RNA can be collected using standard PAXgene™ or Versag ene(TM) (Gentra Systems, Inc., Minneapolis, M. inn.) RNA can be extracted using the Versagene RNA extraction protocol. The PAXgene™ kit provides higher yields of RNA from PAXgene™ RNA tubes. After RNA extraction, whole blood globin reduction was performed using GLOBIN CIea. r™ (Ambion, Austin, Tex.) can be used. The method uses a bead-oligonucleotide construct to bind globin mRNA. In our experiments, we removed more than 90% of the globin mRNA. The removal of abundant and uninteresting transcripts by the technology allows This may increase the sensitivity of the assay, such as with respect to a cross-array platform.

[0079] RNA quality can be assessed by several means. For example, RNA quality can be Immediately after extraction, the samples were evaluated using an Agilent 2100 Bioanalyzer. This analysis yields an RNA integrity number (RNA integrity score) as a quantitative measure of RNA quality. In addition, after globin reduction, the sample is The samples can be compared to globin-reduced standards. In addition, scaling factors and buffer Background can be assessed after hybridization with microarrays. .

[0080] Real-time PCR can be used to rapidly identify gene expression from whole blood samples For example, isolated RNA can be reverse transcribed and then amplified to produce dsDNA fragments. A non-specific fluorescent dye that is intercalated into the probe or hybridizes with its complementary DNA target. Sequence-specific DNA probes labeled with fluorescent reporters that allow detection only after dimerization. The signal can be detected in real time using the lobes.

[0081] Hence, it may be used by the platform in accordance with the methods disclosed herein. It should be understood that there are many methods of mRNA quantification and detection.

[0082] Expression levels are typically determined by measuring the expression of a particular promoter using methods routinely practiced by those of skill in the art. Depending on the platform, detection and quantification are followed by normalization.

[0083] Platform-appropriate methods for mRNA detection and quantification, as well as matched normalization In terms of theory, it is simply a carefully selected and judged patient trial for the training method. For example, the cohorts listed below are Appropriate weightings that can be used with the genes in the three signatures in the classifier for each These subject-samples were also used to derive differential mRNA detection coefficients. and coefficients and cutoffs for tests performed using the quantification platform could be used to derive.

[0084] In some embodiments, the individual categories of the classifier (i.e., bacterial ARI, Viral ARI, non-communicable diseases) is a cohort that encompasses a variety of such causes. For example, a bacterial ARI classifier may be formed from multiple bacterial genera and / or species. The viral ARI classifier was derived from a cohort with multiple bacterial infections and was From cohorts with viral infections from genus and / or species, non-infectious disease classification The children are from a cohort with multiple non-communicable causes of non-communicable diseases, e.g. See Table 8. In this way, each of the resulting classifiers can be classified according to the underlying bacterial, viral, and is blind to non-infectious causes. Some or all of the subjects with a non-infectious etiology in this study have symptoms consistent with respiratory infection. do.

[0085] In some embodiments, the signatures are trained using a selected set of patient samples. Gene sets are identified by the model according to their ability to segregate phenotypes during the process can be obtained using a supervised statistical approach known as sparse linear classification. The outcome of the training is a gene signature and classification coefficients for the three comparisons. Together, the signatures and coefficients provide a classifier or predictor. Training also involves the selection of a threshold or A threshold or cutoff value may be used to establish a threshold or cutoff value for test performance. For example, the sensitivity and specificity of the test can be adjusted to vary. The threshold for bacterial ARIs is set to increase the sensitivity of the test for bacterial infection, if necessary. may be intentionally lowered to

[0086] In some embodiments, the classifier generation comprises: (i) determining whether each normalized gene expression value is a positive integer; The weights for each gene are assigned and the expression values ​​are used to classify the gene into a classifier (e.g., linear Regression classifier) ​​equation and determine a score for each outcome for multiple subjects. (ii) determining the accuracy of classification for each outcome across multiple subjects. and then (iii) adjusting the weights until the classification accuracy is optimized. The method includes iteratively adjusting the weights of genes in each classifier. Can be enjoyed.

[0087] In some embodiments, the classifier is a linear regression classifier, and said generating comprises linking and converting the classifier scores to probabilities using a function. As is known, the link function is the relationship between the target / output of the model (e.g., the probability of bacterial infection) and Identifying the links between systematic components of a linear model (in this case, combinations of explanatory variables, including predictors). It determines how the expected value of the response is related to the linear predictor of the explanatory variables. show.

[0088] Method of classification The present disclosure provides a method for determining whether a patient's respiratory illness is due to a bacterial infection, a viral infection, or The present invention further provides a method for determining whether the infection is due to infectious or non-infectious causes. The method for this relies on the use of classifiers obtained as taught herein. a) a defined set of genes (i.e., a set of genes that are associated with one or more of the three signatures) b) determining the expression level of the gene; c) normalizing gene expression levels for each G; These are then combined with predefined weighting values ​​(coefficients) for each of the genes in each signature. Bacteria classifiers, viral classifiers, and / or non-infectious disease classifiers (i.e. d) inputting a predefined threshold, cutoff value, indicating the likelihood of infection; comparing the output of the classifier with a confidence interval or range of values; and optionally, e) comparing the output of the classifier with a confidence interval or range of values; The method may include reporting the results of the above steps together.

[0089] A simple overview of such a method is provided in Figure 3. In this representation, three genes Each child signature represents a different ARI etiology (bacterial or viral), or infection. These signals do not provide information about the patient's host response to the disease state (NI). Necha can be caused by bacterial ARI, viral ARI, or a non-infectious but diseased state. Consistent and coordinated increases in expression levels in response to one of three clinical conditions or These signatures are carefully determined groups of gene transcripts that cause a decrease in A group of patient samples having the condition of interest is used to guide the study (training 10).

[0090] With reference to FIG. 3, after obtaining a biological sample (e.g., a blood sample) from a patient, several steps are performed. In this form, mRNA is extracted. The quantification is performed for all or a subset of the genes in the signature. Depending on the instrument used, mRNA may first have to be purified from the sample. do.

[0091] The signature is a reflection of the clinical state and is distinct from at least one of the other two possibilities. For example, the bacterial ARI signature is defined as a set of characteristics that distinguish patients with bacterial ARI from those without bacterial ARI. Patients without ARI (viral ARI or non-communicable disease if in this indication) A group of biomarkers (herein, gene mRNA The viral ARI signature is a set of genes that are expressed by the viral transcripts. Patients with viral ARI and those without viral ARI (bacterial ARI or non-infectious disease) It is defined by a group of biomarkers that distinguish between patients with either The infectious disease signature identifies patients with non-infectious etiologies, either bacterial or viral. It is defined by a group of biomarkers that distinguish between patients with either ARI.

[0092] The normalized expression levels of each gene in the signature (e.g., first column of Table 9) were used in the classifier. The explanatory or independent variables or features that are used to classify the It can have the general form as a reduction formula: P(having condition)=Φ(β1X1+β2X2+ …+β d X d )(Formula 1) where the condition is bacterial ARI, viral ARI, or non-infectious illness; Φ(.) is is the probit (or logistic, etc.) link function; {β1, β2, …, β d} are the coefficients obtained during training (e.g., columns 2, 3, and 4 from Table 9) (The coefficients are also referred to herein as “weights”: {w1, w2, …, w d});{ X1,X2,…,X d} is the normalized gene expression level of the signature; and is the size of the signature (i.e., the number of genes).

[0093] As will be understood by those skilled in the art, the coefficient values ​​for each explanatory variable are calculated using a probit regression. Used to measure the expression of genes or a subset of genes used in the regression model Varies by technology platform. For example, Affymetrix U13 Classifier algorithm for gene expression measured by 3A 2.0 microarrays The coefficients for each of the features in are shown in Table 9.

[0094] The sensitivity, specificity, and overall accuracy of each classifier were calculated using receiver operating characteristic (ROC) curves. This can be optimized by varying the threshold for classification.

[0095] Another aspect of the present disclosure is a method for treating an acute respiratory infection (ARI) in a subject, the method comprising administering to the subject an effective amount of the ARI, the ARI being bacterial in origin or viral. 1. A method for determining whether a biological sample is of infectious or non-infectious origin, comprising: a) extracting biological material from a subject; (b) obtaining a gene expression profile of interest from the biological sample; The expression levels of all gene sets (i.e., the three signatures) were evaluated. c) determining gene expression for the technology used to perform said measurements; d) normalizing the levels to produce normalized values ​​for each signature; a bacterial classifier with predefined weighting values ​​(coefficients) for each of the genes in inputting into a virus classifier, and a non-infectious disease classifier (i.e., predictor); e) classifying the output of said classifier into a defined threshold, cutoff value, or range of values ​​that indicate the likelihood of infection. and e) comparing said sample to a range of pathogenicity, such as bacterial etiology, viral etiology, or non-viral etiology. Classifying a disease as an infectious disease, In some embodiments, the method provides a method for determining the probability of ARI etiology. The method further includes generating a report assigning a score to the patient based on the score.

[0096] During training, a classifier developed using the training sample set diagnoses new individuals ("classification"). For each subject or patient, a biological sample is collected and The samples were analyzed to determine the identity of each gene identified by the signatures found during training. The normalized expression levels (i.e., relative mRNA expression levels) in the 100-kDa subunits are the inputs for the classifier. The classifier also uses weighting coefficients discovered during training for each gene. As such, the classifier is used to calculate three probability values. Each probability value is the bacterial AR The prospects for three possible clinical conditions are I, viral ARI, and non-infectious disease. It can be used to determine

[0097] In some embodiments, the results of each of the classifiers, i.e., new subjects or The probability that a patient has a bacterial ARI, a viral ARI, or a non-communicable disease is reported In its final form, the three signatures with corresponding coefficients are applied to each individual patient. Three probability values ​​were generated: bacterial ARI, viral ARI, and non-communicable disease. In some embodiments, these values ​​are the confidence with which a classification is made. In some embodiments, the output of the classifier can be reported relative to a reference range that indicates the degree of , and compared to a threshold to determine, for example, a classifier score or probability for bacterial ARI, viral ARI, A positive result is reported if the RI or threshold indicating the presence of one or more non-communicable diseases is exceeded. If the classifier score or probability fails to reach the threshold, the result may be reported. The results are reported as "negative" for each of the conditions. Only values ​​for infectious ARI are reported, and the reports are for people who are ill but not infected. The Company will not comment on the possibility of this not happening.

[0098] Classifiers obtained on one platform may not perform optimally on another platform. It should be noted that this may not be the case. This may be due to probe hybridization or This may be due to other technical issues specific to one platform. Adapting signatures as taught herein from one platform to another Methods for making the same are also described herein.

[0099] For example, the signatures obtained from the Affymetrix platform include About genes and / or the Affymetrix platform in Necha For surrogate genes that correlated with genes in the signature, the corresponding TLDA profiles were The use of lobes can be adapted to the TLDA platform. trix probes and the genes they measure, plus, for technical reasons, TLDA A platform that does not perform well or for which no cognate TLDA probe exists for gene probes, either to replace ffymetrix probes A list of "replacement genes" that will be introduced as replacements is shown. These replacements are These may indicate highly correlated genes or may associate with different positions in the same gene transcript. The probe may be a pan-viral gene probe. Additional genes may be included, such as a pan-viral gene probe. Table 1 The weights shown in are calculated for classifiers run on the microarray platform. Weights that have not been estimated are indicated in the table by "NA" (see below). Example 4 of this paper provides a complete translation of these classifiers into the TLDA platform. Reference probes for TLDA (i.e., normalization genes, e.g., TRAP 1, PPIB, GAPDH, and 18S) were also used as weights and the Affymetrix profile. In the columns for lobe set IDs (these are not part of the classifier), indicate "NA." Additional genes that do not necessarily correspond to Affymetrix probe sets are included. Gene probes also show "NA" in the Affymetrix Probe Set ID column. will be done.

[0100] JPEG2025069171000002.jpg233146JPEG2025069171000003.jpg240129JPEG2025069171000004.jpg239124JPEG2025069171000005.jpg93133

[0101] Further discussion of this example signature on the TLDA platform can be found in the following implementation: Provided in Examples 3 and 4.

[0102] This method of determining the etiology of ARI may be combined with other tests. For example, if the patient If determined to have a viral ARI, follow-up testing should be for influenza. Whether the virus can be directly detected as virus A or B or whether it is a host that indicates such infection The aim of the present invention may be to determine whether a primary response can be detected. Follow-up tests to I results are directed at detecting gram-positive or gram-negative bacteria. whether they can detect host responses indicative of such infections In some embodiments, a classifier may be used to determine the class of infection. and also to identify specific pathogens using pathogen-specific probes or detection methods. Concurrent testing can be performed to test for the body. No. 2015 / 0284780 (Methods for detecting active tuberculosis); Tsalik et al. See U.S. Patent Publication No. 2014 / 0323391 (Methods for Classifying Bacterial Infections).

[0103] Methods for determining secondary classification of ARI in a subject The present disclosure also provides a method for classifying a subject using a secondary classification scheme. Another aspect of the present invention relates to a method for treating an acute respiratory infection (ARI) in a subject, the method comprising administering to the subject an effective amount of the ARI, the ARI being a bacterial or viral infection of any kind. 1. A method for determining whether a biological sample is of infectious or non-infectious origin, comprising: (a) removing a biological sample from a subject; (b) obtaining a subject's gene expression profile from the biological sample; By assessing the expression levels of predefined gene sets (i.e., the three signatures), (c) determining the necessary information regarding the technology used to perform said measurements; Optionally, normalizing the gene expression levels to generate normalized values; (d) for each signal Classification with predefined weighting values ​​(coefficients) for each of the genes in Necha (e) inputting the normalized value into a classifier (i.e., a predictor); , a step of comparing the sample with a predefined threshold, cutoff value, or range of values ​​that indicate the likelihood of infection. (f) if the sample is negative for bacteria, a viral classifier and a non-infectious repeating step (d) using only the classifier; and (g) filtering said sample. The method includes, and comprises, classifying the disease as having an infectious or non-infectious etiology; or a method consisting essentially of them.

[0104] Another aspect of the present disclosure is a method for determining whether an acute respiratory infection (ARI) in a subject is caused by a bacterial or viral infection. 1. A method for determining whether a biological sample is of infectious or non-infectious origin, comprising: (a) removing a biological sample from a subject; (b) obtaining a subject's gene expression profile from the biological sample; By assessing the expression levels of predefined gene sets (i.e., the three signatures), (c) determining the technology used to make said measurements; (d) normalizing the gene expression levels to generate normalized values ​​for each signature; A classifier with predefined weighting values ​​(coefficients) for each of the genes (i.e. (e) inputting the normalized value into a classifier (predictor); (f) comparing said antibody to a value, cutoff value, or range of values ​​indicative of likelihood of infection; If the sample is negative for viruses, only the bacterial and non-infectious classifiers are used. repeating step (d) for determining whether the sample is a bacterial pathogen or Classifying a disease as a non-communicable disease, Provide a way to become a target.

[0105] Yet another aspect of the present disclosure is a method for determining whether an acute respiratory infection (ARI) in a subject is of bacterial origin or not. 1. A method for determining whether a virus or a non-infectious virus is of viral or non-infectious origin, comprising: (a) administering to a subject (b) obtaining a biological sample from the biological sample; and (b) detecting a gene expression profile of the subject from the biological sample. by assessing the expression levels of predefined gene sets (i.e., the three signatures). (c) determining the technology used to perform said measurements; (d) normalizing the gene expression levels to generate normalized values ​​for each signature; A classifier (i.e., (e) inputting the normalized value into a classifier (i.e., a predictor); comparing the detected number of samples to a threshold, cutoff value, or range of values ​​indicative of likelihood of infection; ) If the sample is negative for non-infectious diseases, a viral classifier and a bacterial classifier repeating step (d) using only the nucleic acid; and (g) isolating said sample. The method includes, comprises, or consists of classifying the pathogen as being of pulmonary or bacterial etiology. The present invention provides a method consisting essentially of the above.

[0106] In some embodiments, the method further comprises assigning the patient a score indicative of the probability of an ARI. The method further includes the step of generating a report that hits the target.

[0107] Classifying a patient's condition using a secondary classification scheme is shown in Figure 4. In this study, the bacterial ARI classifier classified patients with bacterial ARI into those without bacterial ARI. from patients with a non-infectious ARI (which may instead have a viral ARI or a non-infectious etiology). After this, a secondary classification was imposed on patients with nonbacterial ARI to distinguish between viral ARI and noninfectious ARI. The same process of primary and secondary classification can also be used to further distinguish between viruses and In that case, the patient is determined not to have a viral infection. Patients who are subsequently diagnosed with a bacterial ARI are then secondarily classified as having a bacterial ARI or a non-infectious etiology. Similarly, applying a non-communicable disease classifier as a first-line test could help determine whether a patient has such a non-communicable disease. It is determined whether the patient has an infectious disease or, alternatively, symptoms of an infectious cause. The classification step is to determine whether the infectivity is due to a bacterial or viral pathogen. Determine.

[0108] The results from the three primary and three secondary classifications can be analyzed by those skilled in the art using a variety of techniques. These are then summed (e.g., sum, number, or average) to produce actionable reports about the donors. In some embodiments, the second level of classification can be performed using The genes can be some or all of the genes presented in Table 2.

[0109] In such an example, the three classifiers mentioned above (bacteria classifier, virus classifier, and non- The infectious disease classifier is used to perform the first level of classification. For patients with sexually transmitted infections, a secondary classifier distinguishes between those with viral ARI and those with non-communicable illness. (Figure 4, left panel). A new classifier was used to distinguish viral from non-infectious diseases in patients with (Figure 4, center panel), and in the first step, A new classifier for distinguishing viral from bacterial ARI in patients without HIV infection has been developed. used (Figure 4, right panel).

[0110] In this two-stage method, nine probabilities can arise, and these probabilities can be combined in several ways: Each can be used to estimate the probability of bacterial ARI, viral ARI, and non-communicable disease. Two strategies are described herein for reconciling the three sets of predictions: For example, the highest predicted mean probability: the average of all predicted probabilities for bacterial ARI. The mean was calculated and the overall predicted probability of viral ARI and the overall predicted probability of non-communicable diseases were also calculated. The same is true for the observed probabilities: the highest average probability represents the diagnosis.

[0111] Most predictive: Instead of averaging the predicted probabilities for each condition, we Expected incidence of refractory disease (i.e., bacterial ARI, viral ARI, or non-infectious illness) The best case scenario is that all three classification schemes give the same answer (e.g., Bacterial ARI for scheme 1, bacterial ARI for scheme 2, and bacterial ARI for scheme 3 In the worst case, each scheme specifies a different diagnosis, and The method is to produce a tie.

[0112] Using the training set of patient samples described above, the results of stage 1 separation are shown, for example, in Table 3 (column (clinical classification presented in ; diagnostic test prediction presented in ). Got it.

[0113] JPEG2025069171000006.jpg32167

[0114] After stage 2 classification using the highest predicted mean probability strategy, the results are shown in Table 4 (clinical The classification may be similar to that presented in the table (clinical classification; diagnostic test predictions).

[0115] JPEG2025069171000007.jpg31168

[0116] After stage 2 classification using the highest predicted number strategy, the results are shown in Table 5 (clinical classification presented in columns 1 and 2). The present invention may be similar to the diagnostic test predictions presented in the previous section.

[0117] JPEG2025069171000008.jpg33168

[0118] For example, achieve classification as described above and / or as summarized in Table 2. Table 2 shows the results for three separate classification strategies that answer different diagnostic questions. A total of 270 genes collectively comprising three complex classifiers were included. The first is the same as that presented in Example 1 below, BVS (bacterial These probes are listed in Table 9. The results are the same as those used in the previous study, and Table 9 provides the probe / gene weights used in the classification. They also correspond to the genes presented in Table 10.

[0119] The second is called 2L for two-tier or two-stage. This is the hierarchical scheme presented in Figure 4. It is a room.

[0120] The third is similar to the BVS, but also includes a population of healthy controls (also described in Example 1). This group was poorly controlled for non-infection. It has been shown that there are use cases where discrimination from healthy may be clinically important. For example, this may involve serial measurements of signatures that correlate with recovery. Used to identify patients who have been exposed to infectious agents and are presymptomatic versus asymptomatic. In the BVSH scheme, four groups were included in the training cohort: bacterial ARI, Groups with viral ARI, SIRS (non-infectious disease), and healthy groups are represented. These four groups represent four distinct signatures that distinguish each class from all other possibilities. is used to create

[0121] Table 2 Legend: Probe = Affymetrix probe ID BVS = bacterial ARI, viral ARI, and non-infectious illness (with respiratory symptoms) Three classifier models were trained on patients with A value of 0 indicates that the probe is not present in this classification scheme. It indicates non-existence. BVS-BO = Genes or genes included in the bacterial ARI classifier as part of the BVS classification scheme or probe. This classifier distinguishes patients with bacterial ARI from those with other etiologies (viral ARI). or 10) to specifically identify. BVS-VO = This column identifies genes included in the viral ARI classifier. As for BVS-BO, this classifier distinguishes patients with viral ARI from those with other Specific differentiation of etiology (bacterial ARI or non-infectious disease). BVS-SO = Except this column identifies genes included in the non-communicable disease classifier , as for BVS-BO or BVS-VO. This classifier has a non-communicable disease Specifically distinguish patients from those with other etiologies (bacterial or viral ARI). 2L refers to a two-tiered hierarchical classification scheme. A 1 in this column indicates that a specific probe or genes were included in the classification task. This two-tier classification scheme is by itself It consists of three separate step-by-step tasks: the first step applies one-to-one and the second step applies bacterial AR. A given subject may be classified as having an "other" category If the result falls within the range of 0 to 10, a second stage of classification occurs that distinguishes between the remaining possibilities. A first stage for a model that determines whether a given subject has a non-communicable disease; This was followed by SL-BV, which potentially differentiates between bacterial and viral ARI. A 1 in these columns indicates that the gene or probe is not included in the specified classification model. 2L-BO and 2L-VS constitute alternative two-tier classification schemes. 2L-VO and 2L-SB comprise the third model in the two-tier classification scheme. Finally, BVSH includes healthy individuals in the training cohort and therefore is a good candidate for bacterial ARI, A report containing classifiers for viral ARI or healthy state compared with non-communicable disease The dark grey BVSH column refers to any of the classification schemes included in this classification scheme. Identifies the gene or probe. This scheme itself represents a "1" in these columns. BVSH-BO, BVSH with their respective probe / gene compositions indicated in -VO, BVSH-SO, and BVSH-HO.

[0122] Table 2 shows the virus classifier, the bacteria classifier, and the Provides an overview of the use of members of gene sets for non-communicable disease classifiers. "1" indicates the membership of that gene in the classifier.

[0123] JPEG2025069171000009.jpg237135JPEG2025069171000010.jpg234146 JPEG2025069171000011.jpg236142JPEG2025069171000012.jpg235144JPEG2025069171000013.jpg235148 JPEG2025069171000014.jpg234146JPEG2025069171000015.jpg235148 JPEG2025069171000016.jpg236143JPEG2025069171000017.jpg234147JPEG2025069171000018.jpg236147JPEG2025069171000019.jpg235146 JPEG2025069171000020.jpg235127JPEG2025069171000021.jpg234147JPEG2025069171000022.jpg235142JPEG2025069171000023.jpg234145 JPEG2025069171000024.jpg234148JPEG2025069171000025.jpg23584

[0124] Methods of Treating a Subject with an ARI Another aspect of the disclosure is a method for treating an acute respiratory infection (ARI) of unknown etiology in a subject. (a) obtaining a biological sample from a subject; (b) detecting the biological sample; Gene expression profiles of interest can be collected from a predefined set of genes (e.g., one, two, or three or more signatures) (c) the technology used to make the measurements, as appropriate; (d) normalizing the gene expression levels to generate normalized values; A bacterial classifier with predefined weighting values ​​(coefficients) for each of the genes in the The normalized values ​​are input to the infectious disease classifier and the non-infectious disease classifier (i.e., predictor). (e) filtering the output of the classifier based on a predefined threshold, cut-off value, or threshold for determining whether the infection is present; (f) comparing said sample to a range of likely values; (g) classifying the disease as a disease caused by a pathogenic agent or a non-communicable disease; and administering to said subject an appropriate treatment regimen as identified thereby, The present invention provides a method comprising, or consisting essentially of,

[0125] In some embodiments, step (g) determines that the etiology of the ARI is bacterial. In another embodiment, step (g) comprises administering an antibacterial treatment if If the etiology of the ARI is determined to be viral, this involves administering antiviral therapy.

[0126] After the etiology of the subject's ARI has been determined, she may consider treatment, e.g., administering a steroid if the ARI is viral. If it is determined that she has a pulmonary embolism, she may receive antiviral treatment and / or she may receive treatment to delay the progression of the infection. Alternatively, if the ARI is determined to be bacterial, A bacteriological treatment regimen may be administered (e.g., antibiotics). Subjects who are not hospitalized may be sent home or undergo further diagnosis and treatment (e.g., allergies, asthma, The patient may be examined for symptoms such as breathing.

[0127] However, laboratories aim to identify the etiology of ARI and provide appropriate treatment. For example, it is intended that the gene expression levels of the classifiers can be communicated to a physician. The person who performed the test does not need to make a comparison. In addition, the doctor, after examining the patient, A peripheral blood sample is taken, the sample is assayed for the classifier, and the etiologic status of the patient is reported to the agency. The agency is intended to order the agency to report the condition to the physician. The physician shall determine the etiology of the ARI. Upon receipt of a positive result, a physician may order appropriate treatment and / or isolation.

[0128] The methods provided herein allow for proper treatment of patients and reduce inappropriate use of antibiotics. Therefore, the present invention can be effectively used to diagnose the etiology of a disease. The provided methods have a variety of other uses, including: (1) treating a subject that has been exposed to a pathogen; Host-based detection of individuals with impending but non-symptomatic disease Studies based on the nature of the disease (e.g., in a scenario of natural spread of the disease through a population, but (2) in clinical trial settings or in population models of immunity; for monitoring the response to a vaccine or drug, either for host-based testing, (3) for screening for impending disease prior to deployment Host-based testing (e.g., military deployment or civilian populations such as those aboard cruise ships) scenario), and (4) ARIs in livestock (e.g., avian influenza and other pandemic Host-based screening for potentially endemic viruses This includes, but is not limited to, testing.

[0129] Another aspect of the present disclosure is a method for extracting a biological sample comprising: (a) a means for extracting a biological sample; (b) a method for extracting a biological sample; Multiple synthetic oligonucleotides with regions of homology to a group of gene transcripts as shown (c) a means for producing one or more arrays of optides; and (c) instructions for use. A method for treating acute respiratory syndrome in a subject, comprising, consisting of, or consisting essentially of The present invention provides a kit for determining the etiology of an acute respiratory infection (ARI).

[0130] Yet another aspect of the present disclosure is a kit for evaluating acute respiratory infection (ARI) classifiers. (a) detecting a group of gene transcripts as taught herein; One or more arrays of synthetic oligonucleotides having regions of homology to the target gene. (b) preparing a nucleic acid sequence comprising an oligonucleotide having a region homologous to a normalization gene; (c) adding to the array a sample of a subject suffering from an acute respiratory infection (ARI); (d) obtaining a biological sample; (e) isolating RNA from said sample and (e) a step of measuring the transcriptome on the array. (f) normalizing the transcriptome measurements to normalizing genes and classifying electronically transferring the normalized measurements to a computer for executing a sub-algorithm. (g) generating a report, and optionally, (h) taking appropriate action based on said results. The present invention provides a method comprising, consisting of, or consisting essentially of the steps of: do.

[0131] Classification System With reference to FIG. 11, a classification system and / or computer program product 1100 in a platform, or Thereby, it can be used. Classification system and / or computer program product 1100 may include any suitable software, firmware, and / or hardware. Operable to receive, transmit, process, and store data using any combination of and may be standalone and / or part of the global network known as the Internet. any normally public and / or by private, real and / or virtual, wired and / or wireless networks. One or more enterprise, application, personal, pervasive, and and / or embedded computer systems, and may have various types of tangible representations. The present invention may include a non-transitory computer readable medium.

[0132] As shown in FIG. 11, the classification system 1100 includes one or more operating operating system and / or one or more applications running on A processor subsystem 1140 may include multiple central processing units (CPUs). Although one processor 1140 is shown, there may be multiple processors 1140. can be either electrically interconnected or independent. It will be understood that the processor 1140 may receive data from a memory device, such as the memory 1150. Executing the computer program code to perform some of the operations and methods described herein. Any conventional or special purpose processor may be configured to implement at least a portion of the The processor may include a digital signal processor (DSP), a field programmer, FPGAs, Application Specific Integrated Circuits (ASICs), and Multicore This includes, but is not limited to, a processor.

[0133] The memory subsystem 1150 includes random access memory (RAM), read-only memory (ROM), and Memory (ROM), erasable programmable read-only memory (EPROM) or A hierarchy of memory devices such as flash memory and / or any other solid-state memory device. may include.

[0134] A storage circuit 1170 may also be provided, which may include, for example, a portable computer diskette. hard disk, portable compact disk read-only memory (CDROM) , optical storage devices, magnetic storage devices, and / or any other type of disk or tape The memory circuitry 1170 may include a group-based memory subsystem. 00. The memory circuit 11 may provide non-volatile storage of data / parameters / classifiers for 70 may include disk drives and / or network storage components. The memory circuit 1170 may store code and / or applications that may be executed by the processor 1140. In some embodiments, the memory can be used to store data that can be accessed by the user. The memory circuit 1170 stores the data used by the classification system 1110, such as signatures, weights, thresholds, etc. A database providing access to one or more data / parameters / classifiers may be stored. Any combination of computer-readable media may be utilized by the storage circuit 1170. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer readable storage medium may be, for example, an electronic, magnetic, optical, electromagnetic, infrared, wire, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing More specific examples of computer-readable storage media (including, but not limited to, (non-relevant list) includes portable computer diskettes, hard disks, random Random Access Memory (RAM), Read-Only Memory (ROM), Erasable Programmable Read-only memory (EPROM or flash memory), portable compact disks CD-ROM, optical storage device, magnetic storage device, or any of the foregoing As used herein, a computer readable storage medium includes any suitable combination of The body is used by or in connection with an instruction execution system, device, or apparatus. The program may be any tangible medium capable of containing or storing a program.

[0135] The input / output circuitry 1160 may include a display, and / or a keyboard, touch screen, may include a user input device such as a screen and / or a pointing device. The devices attached to the input / output circuitry 1160 may be used by a user of the sorting system 1100. The input / output circuit 1160 may be used to provide information to the processor 1140. The devices included are networking or communication control devices, input devices (keyboard, mouse) , touch screen, etc.), and output device (printer or display). The input / output circuitry 1160 may also include devices such as a display and / or a printer. The classification system 1100 may provide an interface to a device, and the results of the operation of the classification system 1100 may The information may be transmitted to those devices so as to be provided to users of the system 1100.

[0136] Optional update circuitry 1180 provides updates to the classification system 1100. The updates may be included as an interface to the memory 1150 and / or or a code executed by the processor 1140 stored in the memory circuit 1170. The updates provided via the update circuit 1180 may include Also, a database that maintains information about the classification system 1100, such as signatures, weights, thresholds, etc. Upgrading the portion of the storage circuit 1170 associated with a database and / or other data storage format It may include updates.

[0137] The sample input circuit 1110 of the sorting system 1100 is a platform as described above. The platform may provide an interface for receiving a biological sample to be analyzed. The sample input circuit 1110 receives biological samples provided by a user into the classification system 1100. A classification system 1100 and / or platform for receiving and processing biological samples. The sample input circuit 1 may include mechanical and electrical elements that transport the biological sample within the sample chamber. 110 identifies the barcoded container for identification of the sample and / or test order. The sample processing circuit 1120 processes biological samples for automated analysis. To prepare the sample, the classification system 1100 and / or the biological sample in the platform may be used. The sample analysis circuit 1130 may further process the processed biological sample. The sample analysis circuit 1130 may analyze the biological samples provided to the classification system 1100. Regarding the materials, for example, the materials used to measure gene expression levels of a defined set of genes The sample analysis circuit 1130 may also perform normalization by normalizing gene expression levels. The sample analysis circuit 1130 may generate gene expression values. Respiratory infection (ARI) classifier, viral ARI classifier, and non-infectious disease classifier These classifiers can be derived from the defined gene sets for each of the genes. The sample analysis circuit 1130 includes a predefined weighting value (i.e., coefficient). Select from the following classifiers: Arthropoietin Infection (ARI), Viral ARI, and Non-Communicable Disease. The normalized gene expression values ​​may be input into one or more selected acute respiratory disease classifiers. The analysis circuit 1130 may be configured to determine one or more of bacterial ARI, viral ARI, and non-infectious illness. calculates a probability of etiology for a plurality of the subjects based on the classifier, and Deciding whether the infection is of bacterial, viral, or non-infectious origin, or some combination thereof Certain outputs may be controlled via input / output circuitry 1160.

[0138] A sample input circuit 1110, a sample processing circuit 1120, a sample analysis circuit 1130, an input / output circuit The circuit 1160, the memory circuit 1170, and / or the update circuit 1180 may include at least Also partly under the control of one or more processors 1140 of the classification system 1100. As used herein, executing "under the control of" the processor 1140 means Sample input circuit 1110, sample processing circuit 1120, sample analysis circuit 1130, input / output circuit 1160, the storage circuit 1170, and / or the update circuit 1180. The operations are at least partially performed and / or directed by the processor 1140. However, at least some of the operation of those components may be independently controlled, either electrically or mechanically. This does not mean that the processing is automatically performed by the processor 1140. Execution of the program code causes the classification system 11 to operate as described herein. 00's operation.

[0139] Computer program code for performing operations in accordance with aspects of the present disclosure may include Object-oriented programming languages, such as Java, Scala, Smalltalk, k, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python on traditional procedural programming languages, e.g., the "C" programming language, V isual Basic, Fortran 2003, Perl, COBOL 2002 , PHP, ABAP, Dynamic programming languages, e.g. Python, Ruby One or more programs, including .y and Groovy, or other programming languages The program code may be written in any combination of programming languages. Part of the software package is a standalone software package that runs on the classification system 1100. As a cage, partly on the classification system 1100 and partly on a remote computer The latter may be performed locally or entirely on a remote computer or server. In this situation, the remote computer is a computer that is part of a local area network (LAN) or The classification system 110 can be implemented over any type of network, including a wide area network (WAN). 0, or the connection may be to an external computer (through the internet) or to a cloud computing environment or Software as a Service (SaaS) It can be provided as a service such as:

[0140] In some embodiments, the system provides quantitative or semi-quantitative detection of gene expression. , a computer readable code that can be converted into a cumulative score or probability of etiology of ARI. Includes.

[0141] In some embodiments, the system allows the user to select a biological sample to be tested. Then, after some time (preferably a short time, e.g., 30 minutes or 45 minutes, or 1 hour, 2 hours, or 3 hours, up to 8 hours, 12 hours , 24 hours, or 48 hours) to receive the results from the system. The NIH 2010 / 2013 NIH 2013 / 2013 NIH is a sample preparation-to-results system with integrated components, such as:

[0142] The invention relates in its application to the invention as set forth in the following description or as illustrated in the following drawings, It should be understood that no limitations are imposed on the details of construction or arrangement of components. The invention is capable of other embodiments and of being practiced or carried out in various ways. .

[0143] The recitation of ranges of values ​​herein is intended to include the range unless otherwise indicated herein. It is intended simply to serve as a shorthand way of individually referring to each separate value that falls within and each separate value is intended to be treated as if it were individually listed herein. All methods described herein are incorporated herein by reference in their entirety. Unless otherwise indicated in the Detailed Description or otherwise clearly contradicted by the context , may be performed in any suitable order. The use of exemplary language (eg, "such as") may more clearly illustrate the invention. It is merely intended for clarity and, unless otherwise stated, does not constitute a part of the present invention. No language in this specification is intended to imply any limitation on the scope of any claim. No non-existent element should be construed as essential to the practice of the invention.

[0144] Any numerical range recited herein is inclusive of all values ​​from the lower limit to the upper limit. For example, if the concentration range is stated as 1% to 50%, Values ​​such as %~40%, 10%~30%, or 1%~3% may be expressly set forth in this specification. These are merely examples of what is specifically intended and are not intended to be limiting. All possible combinations of values ​​between and including the lower and upper values ​​are contemplated in this application. should be considered as expressly stated in the

[0145] The following examples are illustrative only and are not intended to be limiting in scope. . EXAMPLES

[0146] [Example 1] Host gene expression classifier diagnoses etiology of acute respiratory illness Acute respiratory infections due to bacterial or viral pathogens are the most common hospital-acquired illnesses. This is one of the reasons why current pathogen-based diagnostic approaches are unreliable and The timeliness of antibiotic treatment is not timely and therefore most patients receive inappropriate antibiotics. Host response biomarkers offer an alternative diagnostic approach to direct antimicrobial use. Provides chi.

[0147] We demonstrate that host gene expression patterns can predict infectious causes of disease in acute care settings. Among those with acute respiratory infections, the present invention They determine whether an infectious disease is caused by a viral or bacterial pathogen. It was determined.

[0148] The samples on which the findings were based were collected from four tertiary care hospital emergency departments and a student health facility. The study was drawn from an observational cohort study conducted in 2016. Forty-four healthy controls and community-onset acute A total of 273 patients with acute respiratory infection or non-infectious disease were randomly selected for inclusion in the study, who were suspected to have sepsis. The study was selected from a larger cohort of patients (CAPSOD study). The mean age was 45 years. Of the 116 participants, 45% were male. Further demographic information is incorporated herein by reference. Published in Tsalik et al. (2016) Sci Transl Med 9(3 22):1-9, Table 1.

[0149] Clinical phenotypes were determined by manual chart review. Routine microbiological testing and multiplex PCR were performed. Peripheral whole blood gene expression was analyzed by microarray analysis. Sparse logistic regression was used to distinguish between bacterial vs. viral A classifier for non-communicable diseases was developed using five independently derived The dataset was used for validation.

[0150] The gene expression-based classifier was used for bacterial acute respiratory infections (71 probes), viral infections (100 probes), and avian influenza (100 probes). acute respiratory infection of viral etiology (33 probes), or non-infectious etiology (26 probes) The three classifiers were applied to 273 patients, and the best class assignment was determined. The overall accuracy was 87% (238 / 273 were in agreement with clinical judgment). (78%, p<0.03) and bacterial vs. viral infections. The proposed classifier was more accurate than three published classifiers for rabies infections (78–83%). The developed classifier was externally validated on five publicly available datasets (AU C 0.90-0.99). We have shown that the classification accuracy of the test based on host gene expression is , Procalcitrant, including bacterial or viral acute respiratory infections or non-infectious diseases The results were compared with those of the NIH and clinically adjudicated diagnoses.

[0151] The host peripheral blood gene expression response to infection may provide a complementary diagnostic tool to those already in use. Providing strategies for cutting 8 This strategy is useful for preventing viral ARIs. 8~13 and thin Fungal ARI 11,14 These advances have been successful in characterizing the host response to Nonetheless, their diagnostic use in patient care settings has been challenged by several issues. Important considerations in the development of host-based molecular signatures include: The main problem is that they are developed in the intended population of users. 15 However, However, almost all published gene expression-based ARI classifiers use healthy controls as controls. They use similar individuals and focus on small or homogenous populations, and therefore It is not optimized for use in acute care settings where patients present with undifferentiated symptoms. Additionally, statistical methods used to identify gene expression classifiers include clustering, These strategies often contain overlapping genes based on univariate tests or pathway associations. Although TEZY identifies relevant biology, it does not maximize diagnostic power. An alternative approach, as exemplified, is to combine genes from unrelated pathways to The goal is to create an informative classifier.

[0152] method Classifier-guided cohorts The study was approved by the relevant Institutional Review Board and complied with the Declaration of Helsinki. The subjects or their legally authorized representatives provided written informed consent. Provided.

[0153] Community-acquired pneumonia and sepsis outcomes diagnostic study (clinical trial identification number NCT00258869 ) as part of Duke University Medical Center (D Emergency Department, Durham VA Medical Center, Durham, NC ter (DVAMC; Durham, NC), or Henry Ford Hospi- ter tal (Detroit, MI), patients with suspected community-acquired infections were Registered 16~19 As part of a community-acquired pneumonia and sepsis study, UNC Health Care Emergency Department Additional patients were enrolled through the National Center for Cardiovascular Diseases (UNC; Chapel Hill, NC). suspected infection and 2 or more systemic inflammatory response syndrome (SIRS) Patients were eligible if they presented with diagnostic criteria 20 ARI cases were reported by the Department of Emergency Medicine (SWG) upper or lower respiratory tract as determined by a physician in charge of infectious diseases (EBQ) or infectious diseases (ELT) Patients with gastrointestinal symptoms were included. Assessment was performed at least 28 days after enrollment and based on previously published Prior to any gene expression-based categorization using established diagnostic criteria, a retrospective study was performed. Based on retrospective manual chart review 17 The findings used to support these determinations The full range of information may not have been available to clinicians at the time of their assessment. The study included 4 patients with pulmonary emphysema and 66 patients with pneumonia who had microbiologically confirmed bacterial ARI. Seventy patients with pulmonary edema were identified. The microbiological etiology was determined by routine culture of blood or respiratory samples. culture, using urinary antigen tests (streptococci or legionella), or serological tests (mycobacteria). Patients with viral ARI (n=115) had viral Confirmation was based on identification of the etiology and compatible symptoms. In addition, Duke as part of the DARPA Health Disease Prediction Study with confirmed viral ARI The study included 48 students from a university. The clinical trial will use the ResPlex II v2.0 viral PCR multiplex assay (Qiage The panel was supported by the National Institute of Infectious Diseases (NIID) Hilden, Germany. Types A and B, Adenovirus (B, E), Parainfluenza types 1-4, Respiratory syncytial virus Somatic viruses types A and B, human metapneumovirus, human rhinovirus, coronavirus Virus (229E, OC43, NL63, HKU1), Coxsackie / Echovirus, and bocavirus. The results showed that some of the patients enrolled had non-communicable diseases. It was determined that the diagnosis was 100% correct (n = 88) (Table 8). "Non-infectious disease" was only considered if the results of the microbiological tests presented did not support an infectious etiology. Finally, healthy controls (n = 16) were selected from among asymptomatic healthy volunteers. = 44; median age 30 years; range 23-59 years) was associated with the effect of aspirin on platelet function. Aspirin-induced hypercalcaemia was observed in 113 subjects enrolled as part of a study on carried out 21 .

[0154] Procalcitonin measurement Concentrations were measured at different times during the study, and as a result, different platforms were used based on efficacy. Several serum measurements were performed using the Roche Elecsys 2010 assay. In the case of the Nalyser (Roche Diagnostics, Laval, Canada) Additional serum measurements were performed using the miniVIDAS immunoassay. Blood was analyzed using a ELISA kit (bioMerieux, Durham NC, USA). If serum is not available, B·R·A·H·M·S PCT-sensitive KRYPTOR (The rmo Fisher Scientific, Portage MI, USA) Immunofluorescence in plasma-EDTA was performed using Phadia Immunology Ref The measurements were performed by the ERENCE Laboratory. Several paired sera and Replicate studies were performed on plasma samples and demonstrated equivalence in concentrations. All procalcitonin measurements were treated equivalently regardless of testing platform.

[0155] Microarray fabrication At the first clinical presentation, patients were enrolled and samples were collected for analysis. After performing the above steps, 317 subjects with a clear clinical phenotype were selected for gene expression analysis. Total RNA was prepared using the PAXgene Blood RNA Kit (Qiagen, Vale). RNases were extracted from human blood using a RNase kit (Nicia, CA) according to the manufacturer's protocol. The quantity and quality of A were measured using a Nanodrop spectrometer (Thermo Scientific ific, Waltham, MA) and an Agilent 2100 Bioanalyzer ( The microarrays were evaluated using a 3D chromatograph (Agilent, Santa Clara, CA). Hybridization and data collection were performed using Expression Analysis (Durham, NC) Using the 33A 2.0 array (Affymetrix, Santa Clara, CA), The assay was carried out according to the Affymetrix technical manual.

[0156] statistical analysis Transplantation in 317 subjects (273 disease patients and 44 healthy volunteers) The cryptome was measured in two microarray batches with seven duplicate samples. (GSE63990). Exploratory principal component analysis and hierarchical cluster analysis revealed substantial These were analyzed using a Bayesian fixed effects model to determine the effect of the probe on the The average batch effect was first estimated and removed. They used seven duplicate samples and fitted a robust linear regression model to the Huber loss function. The remaining expression values ​​were then adjusted using a filter.

[0157] Sparse classification methods such as sparse logistic regression run the risk of overfitting. Simultaneous classification and variable selection while reducing 21 Therefore, univariate tests A separate gene selection strategy, such as the sparse factor model or the sparse factor model, is not required. A sparse logistic regression model was run on the 40 probes with the largest variance after batch correction. We fitted each of the binomial tasks separately using 22 Specifically, The authors used a Lasso regularized generalized linear regression with binomial likelihood together with nested cross-validation. The model was used to select the regularization parameters. The code is available in the Glmnet toolbox. It was written in Matlab using the RT-PCR tool. It includes a bacterial ARI classifier, a viral ARI classifier, and a We created a binary classifier for class membership, a non-infectious disease classifier, and a non-infectious disease classifier. Probability of infection (e.g., viral vs. non-infectious in the case of a bacterial ARI classifier) We estimate the maximum membership probability by the condition that We model the three classifiers by following a one-to-many scheme where the rate assigns class labels. These can be combined into a single decision model (called the ARI classifier). 21 . classification ability measurement The criteria were the area under the receiver operating characteristic curve (AUC) for binary outcomes and the mean mean for trinomial outcomes. Confusion matrix for 23 .

[0158] verification The ARI classifier was validated using leave-one-out cross-validation in the same population in which it was derived. A publicly available human gene expression dataset from 328 individuals (GSE626 9, GSE42026, GSE40396, GSE20346, and GSE42834 The data sets were validated independently and externally using at least two clinical trials. Selected if the patient had a bed group (bacterial ARI, viral ARI, or non-infectious disease) To match probes across different microarray platforms, The ARI classifier probes are converted to gene symbols and used to identify the corresponding target microarrays. The array probes were identified.

[0159] result Bacterial ARI classifier, viral ARI classifier, and non-infectious disease classifier In creating a classifier based on host gene expression that distinguishes between clinical conditions, The relevant clinical phenotypes should be represented during the model training process. This is This allows the model to be applied to these included clinical groups, but It will not be possible to apply it to clinical phenotypes that did not exist 15 Targeted collection of ARI diagnosis The group includes patients with viral and bacterial etiology, as well as other possibilities - It must also be distinguished from patients with neither bacterial nor viral ARI. Generally, healthy individuals have served as uninfected controls. However, this does not explain how patients with non-communicable diseases that may have similar clinical symptoms are classified. To the best of our knowledge, this does not take into account the possibility that the In this study, we investigated whether a classifier based on ARI gene expression could be used without the need for disease infection in its derivation. None of them included controls. Therefore, we did not include the subjects with community-onset viral ARI (n= 115), bacterial ARI (n=70), or non-infectious disease (n=88) A large heterogeneous patient population was enrolled at clinical presentation (Table 8). A healthy adult control cohort was used to define the most appropriate control population for ARI classifier development. The study included 44 patients (n=44).

[0160] We first demonstrated that gene expression classifiers derived using healthy individuals as controls were consistent with non- We determined whether we could accurately classify patients with infectious diseases into bacterial ARI, viral ARI, and Using array data from patients with idiopathic ARI and healthy controls, we investigated these conditions We developed a gene expression classifier for . The overall accuracy was 90% and The ARIs were bacterial (AUC 0.96), viral (AUC 0.95), and Highly accurate discrimination between clinically significant and healthy (AUC 1.0) subjects was revealed (Figure 7). However, the classifier can distinguish between ill-uninfected ted) patients, 48 / 88 were bacterial, 35 / 88 were viral, and 5 / 88 were identified as healthy, which is a significant contribution to the biomarker discovery process. This highlights the fact that HIV is not a substitute for patients with non-communicable diseases.

[0161] As a result, we used non-infectious disease controls rather than healthy controls to estimate the incidence of ARI. A classifier was again derived. Specifically, using the array data from these three groups, Three genes in the host response to bacterial ARI, viral ARI, and non-infectious disease An expression classifier was created (Figure 5). Specifically, the bacterial ARI classifier was compared with the viral ARI classifier. The task of unambiguously identifying individuals with bacterial ARIs, either for infectious or non-infectious diseases, is The viral ARI classifier is a classification of viral infections against bacterial ARIs or non-infectious diseases. The non-communicable disease classifier was tasked with unambiguously identifying those with infectious ARI. The intent is to clearly identify all non-communicable diseases requiring adequate labelling of such cases. Not produced in the figure.

[0162] Rather, it was created as an alternative category, and therefore includes both bacterial and viral ARIs. Furthermore, the inventors have determined that healthy people can be assigned to the Since such a disease is unlikely to be a target for such a classification task, It was hypothesized that unstained patients would be more clinically relevant controls.

[0163] Linear support vector machines, supervised factor models, sparse multinomial logistic regression , elastic net, K nearest neighbors, and random forest We developed these gene expression classifiers using six holistic strategies: All performed similarly, but sparse logistic regression used the fewest number of classifiers. It required genes and outperformed other strategies by a small margin (data not shown). We also demonstrated a strategy that created three separate binary classifiers for a given pair. The results were compared with a single multinomial classifier that simultaneously assigned elephants to one of three clinical categories. The latter approach required more genes and was less accurate. We applied a sparse logistic regression model to find 71, 33, and The bacterial ARI classifier, viral ARI classifier, and 26 probe signatures are included in the We defined the RI classifier and the non-communicable disease classifier. The probes and classifier weights are shown in Table 9. As shown.

[0164] Clinical decision making is rarely binary, requiring the simultaneous distinction between multiple diagnostic possibilities. We assigned probabilities of bacterial ARI, viral ARI, and non-infectious illness. We used leave-one-out cross-validation to measure the mean mean of all three, collectively defined as the ARI classifier. A classifier was applied (Figure 6). These conditions are not mutually exclusive. For example, bacterial AR The presence of I does not prevent concurrent viral ARI or non-infectious disease. Furthermore, the assigned probability indicates whether the patient's gene expression response matches the canonical signature of the condition. Each signature represents the degree to which a program matches a given set of rules. Each signature is intended to function independently of the others. In order to simplify the classification, the probabilities are not intended to sum to one. High prediction probability determined class assignment. Overall classification accuracy was 87% (238 / 273 (which was consistent with the determined phenotype).

[0165] Bacterial ARI was identified in 58 / 70 (83%) patients without bacterial infection 179 / 191 (94%) were excluded. Viral ARI was excluded in 90% (104 / 115). and excluded in 92% (145 / 158) of cases. Infection was excluded in 86% (76 / 88) of cases. Considering that the number of different factors, including the location and time of the data, may vary, we performed a positive analysis of all three classifiers. Sensitivity analyses were performed for the positive and negative predictive values ​​(Figure 8). For both classes, the predictive values ​​were based on a wide range of prevalences, including those typically found for ARIs. Rate estimates remained high across the board.

[0166] To determine whether there was any age effect, we used a classification scheme We included it as a variable in the hypothesis that the hypothesis of This is probably due to the large proportion of young people in the viral ARI cohort. However, we found that the correctly classified subjects and the incorrectly classified subjects were No statistically significant differences were observed between the classified subjects (Wilcoxon rank sum p = 0.1 7).

[0167] We compare this performance with widely used biomarkers specific for bacterial infection. Procalcitonin concentrations were compared with those of 23 subjects for whom samples were available. Eight subjects were included and the ARI classifier performance was compared for this subgroup. Procalcitonin concentrations greater than 0.25 μg / L classified patients as having bacterial ARI. whereas values ​​below 0.25 μg / L would assign the patient to either a viral ARI or a non-communicable disease. Procalcitonin was classified as either nonbacterial or possible nonbacterial using the ARI classifier. 186 (78%) of 238 patients were correct compared with 204 / 238 (86%) in the However, the accuracy of the two strategies was The results varied depending on the classification task. For example, distinguishing viral from bacterial ARIs Procalcitonin performed similarly in terms of the 140 / 1400 for the ARI classifier. Compared with 155, 136 / 155 (AUC 0.89) were correctly classified (Yates supplementary Using McNemar's test with a positive p-value = 0.65). However, the ARI classifier distinguishing bacterial ARI from non-infectious illnesses [105 / 124 vs. 79 / 124 (A UC 0.72); p value < 0.001], as well as bacterial ARI compared with viral and non-viral ARI. Distinguishing from all other etiologies, including infectious etiologies [215 / 238 vs. 186 / 238 (A UC 0.82); p-value = 0.02] was significantly superior to procalcitonin. .

[0168] We next compared the ARI classifier to three published genetic studies of bacterial versus viral infections. The descendants were compared to expression classifiers (each of which was derived without including uninfected disease controls). These were 35 probes derived from children with influenza or bacterial sepsis. Classifier (Ramilo) 11 from children with a febrile viral illness or bacterial infection Led 33 probe classifier (Hu) 14 and community-acquired pneumonia or influenza A 29-probe classifier derived from adult ICU patients (Parnell) 12 Contains We used a classifier generated exclusively using patients with viral or bacterial infections. was applied to clinically relevant populations, including patients who were ill but not infected. Specifically, they hypothesized that it would not work well in cases where the immune system is infected with either bacteria or viruses. For individuals who do not have Therefore, we cannot assign the derived We applied the classifiers, as well as published classifiers, to our cohort of 273 patients. The differentiation between infectious, viral, and non-infectious diseases is based on the derived ARI classification. was superior for the classifiers (McNemar's test with Yates correction, Ramiro p=0.002 for Parnell; p=0.0001 for Hu p=0.08) (Table 6) 24,25 This should include non-communicable diseases in the case of ARI. The importance of deriving gene expression classifiers in cohorts representative of the intended use population This highlights the importance 15 .

[0169] Classification of discrepancies To better understand the ARI classifier performance, we analyzed 35 discordant cases. Nine adjudicated bacterial infections were classified as viral and three were classified as noninfectious. Four viral infections were classified as bacterial, and seven were classified as non-infectious. Eight non-infectious cases were classified as bacterial and four as viral. They did not observe a consistent pattern among discordant cases, but notable examples include Included in the nonspecific bacterial infections were 1 patient with M. pneumoniae based on serological conversion 1 of 3 patients with Legionnaires' pneumonia and 1 of 3 patients with viral ARI Cysts in six patients with non-infectious diseases due to autoimmune or inflammatory diseases Of these, only one patient who was determined to have Still's disease was classified as having a bacterial infection. See Tsalik et al. (2016) Sci Tran See also eTable 3 in sl Med 9(322):1-9.

[0170] External Validation Creating classifiers from high-dimensional gene expression data can result in overfitting. Therefore, we compared the five available datasets (GSE6269, GS E42026, GSE40396, GSE20346, and GSE42834) We performed an in silico ARI classifier using gene expression data from 328 individuals represented by These differed in terms of age, geographic distribution, and disease severity. The ARI classifiers were selected because they included at least two relevant clinical groups (Table 7). When applied to four datasets with bacterial and viral ARIs, the AUC The mean mean pneumonia incidence rate (P = 0.90–0.99) was 0.90–0.99. Finally, GSE42834 was significantly lower in patients with bacterial pneumonia (n = 0.90–0.99). 19), including patients with lung cancer (n = 16), and sarcoidosis (n = 68). The overall classification accuracy was 96% (99 / 103), corresponding to an AUC of 0.99. GSE42834 included five subjects with pre- and post-treatment bacterial pneumonia. All demonstrated treatment-dependent resolution of the bacterial infection. Tsalik et al. (2016) Sci Transl Med 9(322):1 Please also see eFigures 3-8 in -9.

[0171] Biological pathways The sparse logistic regression model that produced the classifier is of no additional diagnostic value. , which penalizes the selection of genes from a given pathway. Moreover, such conventional gene enrichment pathway analysis is not feasible. The analysis includes 9,12,14,28,29 Instead, all classifiers are genetic A literature review was conducted on bacterial, viral, and non-viral classifiers (Table 10). The overlap between infectious disease classifiers is shown in Figure 9.

[0172] The viral classifiers include those involved in interferon response, T cell signaling, and RNA processing. The viral classifier included known antiviral response categories such as HIV-1, HIV-2, and HIV-1-associated inflammatory cytokines. and is selected by the virus for the delivery of viral proteins and genomes. Best represents RNA processing pathways such as PNB1 26,27 Its downward control Studies have suggested that it may play an antiviral role in the host response.

[0173] The bacterial classifier encompasses the broadest range of cellular processes, particularly cell cycle control, cell proliferation, and differentiation. Bacterial classifiers identified genes important in T cell, B cell, and NK cell signaling. Unique to the bacterial classifier was the expression of oxidative stress, consistent with sepsis-associated metabolic perturbations. , and genes involved in fatty acid and amino acid metabolism. 28 .

[0174] Summary of clinical indications The present inventors have demonstrated that host gene expression changes are highly specific to the causative pathogen class, and They determined that the pulmonary circulation may be used to identify common causes of respiratory illness. Novel diagnostic platforms to combat unnecessary antibiotic use and emerging antibiotic resistance This creates an opportunity to develop and utilize gene expression classifiers as a form of sparse logistic Using color regression, we identified bacterial etiology and We developed host gene expression profiles that accurately distinguish viral pathogenesis (externally validated AU C 0.90-0.99). Using a non-infectious disease control group to derive an ARI classifier has a wide It yielded high negative predictive value across a wide range of prevalence estimates.

[0175] Respiratory tract infections caused 3.2 million deaths worldwide in 2011, more than any other cause. Resulting in deaths and 164 million disability-adjusted life years lost 1,2 The majority of cases Although most cases are of viral etiology, 73% of outpatients in the US were prescribed antibiotics, accounting for 10% of all antibiotics prescribed in this setting. Accounting for 41% 3,4 Even when a viral pathogen has been microbiologically confirmed, This does not exclude the possibility of simultaneous bacterial infection, but is a "just in case" antibiotic. This empiricism leads to antimicrobial resistance. 5,6 , it is a matter of national security It is recognized as a priority 7 The potential metrics provided in this example are The opportunity to optimize and deliver clinically actionable results that mitigate the emergence of antibiotic resistance Bring about a meeting.

[0176] Several studies have been noteworthy in developing host response diagnostics for ARIs. This was a respiratory virus. 8,10~12,14 , in the ICU population bacterial etiology 12,30 , and tuberculosis 31~33 Typically, these defines the host response profile relative to the healthy state and provides valuable insight into host biology. provide 16,34,35 However, these gene lists are not yet fully understood for diagnostic applications. Suboptimal, but only because the gene expression profile that is the diagnostic component of the test Because it is not representative of the population to which it would apply 15 Healthy individuals are susceptible to acute respiratory disease. They do not exhibit any disease and therefore are excluded from the host response diagnostic development reported herein. will be removed.

[0177] The inclusion of patients with bacterial and viral infections allows for a clear distinction between these two conditions. It allows for the classification of non-communicable diseases, but does not address how to classify non-communicable diseases. It is important to include this phenotype as it indicates possible infectious and non-infectious etiologies. That is, symptoms may not provide the physician with a high degree of diagnostic certainty. A book that uniquely understands the need to include the three most likely states of The approach is to apply such trials to the most deprived and undifferentiated clinical populations. It is possible.

[0178] The few discordant classifications that occurred were due to either misclassification or errors in clinical phenotyping. Errors in clinical phenotyping may have arisen from the current limitations of microbiological diagnosis. This may result from the inability to identify the causative agent. Alternatively, some non-infectious The disease process may in fact be infection-related, by mechanisms yet to be discovered. The cases were not clearly described by a unifying variable such as pathogen type, syndrome, or patient characteristics. Therefore, the gene expression classifiers presented herein can be used to assess the effect of patient-specific variables (e.g., treatment, comorbidities, duration of disease), study-specific variables (e.g., sample processing, assay conditions, RN A quality and yield), or by other factors, including variables yet to be identified. There is a possibility that it may be affected.

[0179] [Example 2] In patients with coinfections defined by identification of bacterial and viral pathogens Classification ability in In addition to determining that age did not significantly affect classification accuracy, we assessed whether the severity of the disease or the etiology of SIRS influenced the classification. Patients with advanced ARI are less likely to suffer from illness as evidenced by lower hospitalization rates. In various cohorts, hospitalization was used as a marker of disease severity. used and assessed its effect on classification ability. The study revealed no difference. In addition, the SIRS control cohort had a respiratory etiology The present inventors included subjects with both respiratory and non-respiratory etiologies. To assess whether classification differs in subjects with S versus non-respiratory SIRS , which was determined to be not different (Fisher exact test p-value of 0.1305).

[0180] Some patients with ARIs have both bacterial and viral pathogens, often called coinfections. However, it is unclear how the host responds to such conditions. It is unclear whether the disease will respond to It can be controlled by bacteria, viruses, both, or neither. The inventors have demonstrated that the bacterial ARI classifier and the viral ARI classifier are the same for bacteria and viruses. The researchers determined how GSE60244 functions in populations with bacterial infections. pneumonia (n=22), viral respiratory tract infection (n=71), and bacterial / viral syncytial infection (n=22). The co-identification group included those with a concurrent diagnosis (n=25) regarding the likelihood of bacterial or viral disease. The presence of both bacterial and viral pathogens was considered without further subcategorization. The present inventors have defined GSE60244 patients with bacterial or viral infections. The classifier was trained on subjects in and then validated on subjects with concurrent identification. (FIG. 10). A host response was considered positive above a probability threshold of 0.5. All four possible categories were observed. Six of the 25 subjects had positive bacterial strains. 14 / 25 had a viral response, and 3 / 25 had positive bacterial and viral responses. 1 / 25 had a virulent signature and 2 / 25 showed none.

[0181] An important clinical decision facing physicians is whether to prescribe an antibacterial agent. The cutoff strategy was based only on the probability of bacterial ARI according to the results from the bacterial ARI classifier. However, information on alternative possibilities, viral or non-infectious, may be available. For example, the probability of an alternative diagnosis is higher than the probability of a bacterial ARI. can have greater confidence in withholding antibacterial agents in patients with low Furthermore, complete diagnostic reporting can identify co-occurring diseases that a single classifier would miss. The inventors have demonstrated that when tested in a population that includes simultaneous identification of bacteria and viruses, These patients are more commonly referred to as "co-infected." To cause infection, there must be a pathogen, a host, and a maladaptive interaction between the two. Simply identifying bacterial and viral pathogens is not enough. The inventors have not yet determined whether a test with evidence of bacterial / viral co-identification exists. Although we cannot know the true infection status of the 25 subjects included, the host response classifier This suggests the existence of multiple host response states. Figure 10 is an informative diagram of the infection state. , can be used by clinicians to diagnose the etiology of ARI.

[0182] References JPEG2025069171000026.jpg225147 JPEG2025069171000027.jpg231151 JPEG2025069171000028.jpg145149

[0183] JPEG2025069171000029.jpg93162

[0184] JPEG2025069171000030.jpg89162

[0185] JPEG2025069171000031.jpg234147 JPEG2025069171000032.jpg167149

[0186] JPEG2025069171000033.jpg235131 JPEG2025069171000034.jpg234143 JPEG2025069171000035.jpg235147JPEG2025069171000036.jpg235147 JPEG2025069171000037.jpg235145JPEG2025069171000038.jpg23554

[0187] JPEG2025069171000039.jpg194155

[0188] [Example 3] Bacterial / viral / SIRS assays designed for the TLDA platform We used a 384-well TaqMan low-density array (TLDA, Applied Custom multi-analyte quantitative real-time analysis on the Biosystems platform The TLDA card is a data Along with multiple endogenous control RNA targets (primer / probe sets) for data normalization, In each well, one or more T For each patient sample, purified The total RNA was reverse transcribed into cDNA and loaded into master wells for microfluidic analysis. Centrifugation through the channel distributes the TaqMan to each assay well. The probes are double-stranded during hybridization with complementary target sequences in each amplification round. 5' to 3' cleavage of the labeled probe resulting in fluorescent signal generation In this manner, accumulated DNA in "real time" is Quantitative detection of PCR products is possible. During exponential amplification and detection, the fluorescent signal The number of PCR cycles at which the detection threshold is exceeded is determined using commercially available software for RT-PCR instruments. Threshold cycle (C t ) or quantification cycle (C q ) To quantify gene expression, the C t However, the intrinsic regularization R C of NA (or the geometric mean of multiple normalized RNAs) t is subtracted from the sample δC for each RNA target in t A value is given that represents the input sample RNA or cDNA Relative expression of target RNA normalized for variations in quantity or quality of A is shown.

[0189] The quantified gene signature data is then computed using the above The signatures are processed according to the probit classifier (Equation 1) described above and reproduced here. The normalized gene expression level of each gene is the explanatory or independent variable used in the classifier. are the features, and in this example, the general form of the classifier is the following probit regression equation: Is: P(having condition)=Φ(β1X1+β2X2+…+β d X d )(Equation 1) where the condition is bacterial ARI, viral ARI, or non-infectious illness; Φ(.) is the probit link function; {β1,β2,…,β d} are the coefficients obtained during training. ;{X1,X2,…,X d} is the normalized gene expression value of the signature; d is the The coefficient values ​​for each explanatory variable are calculated using the probit regression model. The techniques used to measure the expression of genes or subsets of genes used in the model The computer program is specific to the technology platform. The sensitivity, specificity, and overall accuracy of each classifier are calculated. The degree of severity was evaluated by varying the threshold for classification using a receiver operating characteristic (ROC) curve. It is optimized.

[0190] Signatures from the Affymetrix platform (Affy signatures) and Gene identification for TLDA platform based on signatures from other sources A preliminary list is provided in Table 1A below. Appropriate weights for the platform are then determined as described in Example 4 below. It was determined.

[0191] JPEG2025069171000040.jpg232151JPEG2025069171000041.jpg240148JPEG2025069171000042.jpg221148

[0192] [Example 4] The genetic features measured by RT-qPCR proposed in the TLDA platform Bacteria / Virus / SIRS Classification Using Gene Expression The three signature genes constituting the host response-ARI (HR-ARI) test were Custom made by ermoFisher Scientific (Waltham, MA) These gene signatures were then transferred to TaqMan® low-density array cards. Expression was measured using 384-well TaqMan low density arrays (TLDA; Thermo-Fisher) r) platform for custom multi-analyte quantitative real-time PCR (RT-qPCR) PCR) assay. TLDA cards were normalized for RNA loading. Multiple intrinsic parameters used to measure the variability of the Control RNA targets (TRAP1, PPIB, GAPDH, FPGS, DECR1, and 1 8S) along with one or more TaqMan primer / probe sets per well Each primer / probe set was designed and manufactured using the ARI signature. In practice, the normalization procedure is carried out across samples. Select the two reference genes (out of the five available) with the smallest coefficient of variation based on the Primer / probe sets with more than 3% missing values ​​(below the limit of quantification) were discarded. . The remaining missing values ​​(if any) are calculated as 1+max(C q )(C q About RT-qPCR The normalized expression value is then calculated for any given primer / Observed C for the probe set q Calculate as the average of the selected reference minus the Hellemans et al. (2007) Genome Biol 2007;8(2 ):See R19.

[0193] A total of 174 unique primer / probe sets were assayed per sample. Of the primer / probes, 144 primer / probe sets were identified for three AR I-gene signatures of 132 previously described Affymetrix (microarray ) probes representing gene targets (i.e., bacterial gene expression signatures, viral gene expression The present signature, as well as genes in the non-infectious gene expression signature, were measured and six The probe sets are for reference genes, and the inventors additionally identified previously discovered 24 probe sets from the pan-viral gene signature were assayed. No. 8,821,876; Zaas et al., Cell Host Microbe (2009 ) 6(3):207-217. In addition, several primers for "replacement" genes marker / probe sets (these gene expressions are based on the Affymetrix signatures) We added the 1st gene (which is correlated with the expression of some genes) for training. However, when performed with TLDA probes, these genes This is because the RT-qPCR assay on the offspring did not work well.

[0194] For each sample, total RNA was transferred to a PAXgene blood RNA tube (PreAnalyte ix) and purified using the Superscript VILO cDNA synthesis kit (Ther cDNA was extracted using a ELISA kit (http: / / www.ips.go.jp / cDNA ... A standard amount of cDNA for each sample was loaded per master well and then The antibodies are then distributed to each TaqMan assay well via centrifugation through a chromatic fluidic channel. The TaqMan hydrolysis probes hybridize to complementary target sequences in each round of amplification. During dimerization, the dual-labeled probe is cleaved, resulting in the generation of a fluorescent signal. The quantitative detection of fluorescence is called "ligand cleavage." The “real-time” sequence shows accumulated PCR products. During exponential amplification and detection, the fluorescent signal The number of PCR cycles at which the signal crosses the detection threshold is determined by the market Threshold cycle (C) as determined by commercially available software t ) or quantification cycle ( C q).

[0195] Sample / Cohort Selection: Under an IRB-approved protocol, we conducted a randomized controlled trial of 10 patients undergoing emergency department care with acute respiratory illness. Patients in this cohort were enrolled (see Table 11 below). Tsalik et al. (2016) Sci Transl M, incorporated herein by reference. This is a subset of the patients reported in Table 1 of ed 9(322):1-9. Retrospective clinical adjudication of clinical and other test data on bacterial ARI, viral This leads to one of three allocations: infectious ARI, or non-communicable disease.

[0196] JPEG2025069171000043.jpg94154

[0197] Data analysis methodology: During the data pre-processing stage, we aimed to obtain data with minimal coefficients of variation across samples and plates. Select a subset of at least two reference gene targets (out of the five available) We found that targets with more than 33% missing data (17 targets below the limit of quantification) These values ​​are based on the proportion of any particular class, e.g., bacterial ARI. We then discarded the remaining missing values ​​only if they were greater than 1+max(C q ) Then, expression values ​​were normalized for all targets using the previously selected reference combination. , we use the normalized expression value as the C q Selected reference minus the value ( DECR1 and PPIB).

[0198] Once the data were normalized, we used a sparse logistic regression model. Proceed to build a classification model by fitting the model to the data. (Friedman et al. (2010) J. Stat. Softw. 33, 1- 22) This model estimates the probability that a subject belongs to a particular class as a weighted sum of normalized gene targets. Specifically, we estimate p(object belongs to class) = σ(w1x 1+…+w p x p ), where σ is the logistic function and w1, …, w p are the classification weights estimated during the fitting procedure, and x1, …,x p is the normalized expression value represents p gene targets containing

[0199] Similar to the array-based classifier, we compared (1) bacterial ARIs versus viral ARIs; (2) viral ARIs versus bacterial ARIs and non-infectious diseases; and (3) construct three binary classifiers for non-infectious illnesses versus bacterial and viral ARIs. After fitting the three classifiers, we determined p(bacterial ARI), We obtain estimates for p (viral ARI) and p (non-infectious disease). The threshold for each is selected from a receiver operating characteristic (ROC) curve using a symmetric cost function. (expected sensitivity and specificity are approximately equal) (Fawcett (2006) Pattern Recogn Lett 27:861-874). As a result, Elephant is p(bacterial ARI)>t b (t b is the bacterial ARI classifier threshold, Similarly, we used the viral ARI classifier and non-infectious ARI classifier. The thresholds for the sexually transmitted disease classifiers, respectively, tv and t n Select the most appropriate The most likely condition, i.e., the condition with the greatest probability (specifically, is argmax{p(bacterial ARI), p(viral ARI), p(non-infectious disease)} By selecting the

[0200] result: First application of microarray-discovered genomic classifiers to the TLDA platform During the conversion, we assayed 32 samples that were also assayed by microarray. This group performed a TLDA of the gene transcripts that make up the ARI classifier. RT-qPCR measurements were obtained for microarray-based measurements of gene transcripts. The authors replicated previously reported results and therefore classified patients as having bacterial or viral ARI. Ensure that the methodology for classifying individuals as having a non-communicable disease or as having a non-communicable disease is valid. The inventors have developed a TLDA platform and a microarray platform. From the 32 samples tested in both platforms, their corresponding classifiers were used. When assessed by the ESRD, there was an 84.4% agreement, which was due to 27 of 32 subjects The same results were obtained for both the microarray-based and TLDA-based classification models. This means that the results showed the same integrated prediction.

[0201] After demonstrating agreement between microarray-based and TLDA-based classifications, We have previously characterized gene expression patterns in patients with clinically determined ARI status. An additional 63 samples from patients without IL-1 were tested using TLDA-based classification. Thus, a total of 95 samples were evaluated using the TLDA-based classification test. This dataset from 95 samples allowed the development of a TLDA-based RT-qPCR protease inhibitor. How the rat classifies new patients, with only clinical judgement as the reference standard In this experiment, the inventors found that 77 / 95 We observed an overall accuracy of 81.1%, corresponding to correctly classified samples. More specifically, The models achieved 80% (24 out of 30 correct) and 77.4% (31 correct), respectively. 85.3% (24 of 34 correct) and 85.3% (29 of 34 correct) of bacterial ARI accuracy, viral ARI accuracy, and non-communicable disease accuracy were generated. In terms of performance, we compared bacterial ARI classifiers, viral ARI classifiers, and non- The infectious disease classifier had ROC curves of 0.92, 0.86, and 0.91, respectively. The area under the line was observed. However, we did not count the number of cases, and we still did not find any significant difference in the classification ability (accuracy and area under the ROC curve). If partial estimates are desired, we report leave-one-out cross-validated performance metrics. is doing.

[0202] For each of the classifiers (bacterial ARI, viral ARI, and non-infectious diseases), The weights and thresholds are shown in Table 12 below. This table shows the weights and thresholds for 174 genes. Note that instead of targets, we list 151 gene targets, which are listed above. Reference genes were removed in the preprocessing step and missing values ​​were identified as described17. These 17 targets were also removed during the pretreatment step. Removed.

[0203] When the pan-viral signature genes were removed, the AUC, accuracy, and concordance values ​​were significantly higher. There is a slight performance decrease of at most 5% across the percentages.

[0204] summary: The combined host response ARI classifier distinguishes between viral ARIs, bacterial ARIs, and non-infectious diseases. Gene expression signatures, diagnostic of bacterial acute respiratory infections, and a mathematical classification framework for The mathematical classifiers are composed of a set of parameters to determine whether the subject is a bacterial ARI, a viral ARI, or a non- It provides three separate probabilities of having an infectious disease. In each case, the cutoff A threshold value may be specified above which a patient is determined to have the condition. In addition, the threshold value may be altered to change the sensitivity and specificity of the test.

[0205] Measurement of these gene expression levels can occur on a variety of technological platforms. Here, we used a TLDA-based RT-qPCR platform to We also describe the measurement of these signatures. The team will provide platform-specific training to adapt transcript measurement methods (i.e. That is, platform-specific weights, w1, …, w p By establishing Similar direct methodologies can be used to translate gene signatures into other gene expression platforms. The data can then be converted to a platform and subsequently implemented to train the associated classifier. This example also illustrates the TLDA-based classification of ARI etiology and the microarray-based Finally, we demonstrate excellent agreement between the classification of patients with acute respiratory illness. A TLDA-based RT-qPCR platform for diagnosing new patients with and the use of related mathematical classifiers.

[0206] JPEG2025069171000044.jpg236132 JPEG2025069171000045.jpg234146 JPEG2025069171000046.jpg235139JPEG2025069171000047.jpg235149JPEG2025069171000048.jpg235146JPEG2025069171000049.jpg235142 JPEG2025069171000050.jpg234143JPEG2025069171000051.jpg235141JPEG2025069171000052.jpg235146 JPEG2025069171000053.jpg235105

[0207] Any patents or publications mentioned in this specification are indicative of the level of those skilled in the art. Patents and publications are not to be construed as if each individual publication were incorporated by reference. No. 6,393,311 is incorporated herein by reference to the same extent as if each such reference were specifically and individually indicated. In case of conflict, the present specification, including definitions, will control.

[0208] The present invention is thus formulated to accomplish the objects and objects mentioned, as well as the objects and advantages inherent therein. Those skilled in the art will readily recognize that the method described herein can be adequately adapted to obtain the above-mentioned results. The disclosure set forth above represents the presently preferred embodiments and is exemplary and is not intended to be limiting of the scope of the present invention. It is not intended as a limitation of the present invention. Modifications and other uses thereof will occur to those skilled in the art. However, they are within the spirit of the invention as defined by the claims. Included.

Claims

1. 1. A method for generating an acute respiratory illness classifier for a platform, comprising: The classifier may be a bacterial ARI classifier, a viral ARI classifier, or a combination thereof for the platform. and a non-infectious disease classifier, said method comprising: (a) obtaining biological samples from subjects known to be suffering from acute bacterial respiratory infections; obtaining a sample; (b) isolate organisms from multiple subjects known to be suffering from acute viral respiratory infections; obtaining a biological sample; (c) obtaining biological samples from multiple subjects known to have a non-communicable disease; Step, (d) in each of said biological samples from steps (a), (b), and (c), to multiple genes (e.g., all expressed genes or the transcriptome, or measuring gene expression levels of a subset of the genes in the platform; (e) normalizing the gene expression levels obtained in step (d) to obtain normalized gene expression levels; generating a current value; and (f) a bacterial ARI for the platform based on the normalized gene expression values. Creating a classifier, a viral ARI classifier, and a non-infectious illness classifier thereby generating said acute respiratory illness classifier for said platform. How to do it.

2. The measuring step may include purifying cells from the sample, destroying the cells of the sample, and isolating RNA from said sample. The method of claim 1 , wherein the method further comprises or is preceded by the step of:

3. The measuring step may be performed by semi-quantitative PCR and / or nucleic acid probe hybridization. The method of claim 1 or claim 2, further comprising the step of:

4. The platform may be an array platform, a thermal cycler platform, or the like. PCR platforms (e.g., multiplexed and / or real-time PCR platforms), hybridization Chromatinization and multi-signal coding (e.g., fluorescence) detector platforms , a nucleic acid mass spectrometry platform, a nucleic acid sequencing platform, or The method of claim 1 or claim 2, comprising a combination of:

5. The generating step comprises: To provide a classifier for infectious ARI and non-communicable diseases (i) assigning a weight for each normalized gene expression value, and determining the weight for each gene, The expression values ​​are input into a classifier (e.g., a linear regression classifier) ​​equation to obtain a classifier expression value for each of the plurality of subjects. determining a score for the result for (ii) determining the accuracy of classification for each outcome across the plurality of subjects; and And then (iii) adjusting the weights until the classification accuracy is optimized; repeatedly, where genes with non-zero weights are included in each classifier, and optionally, each The components of the classifier (genes, weights, and / or etiology thresholds) are compared against one or more databases. The method of claim 1 , further comprising the step of uploading the file onto a storage medium.

6. The classifier is a linear regression classifier, and the generating step is a step of generating a score of the classifier based on a probability.

6. The method of claim 5, comprising converting

7. The ARI classifier is compared to a known dataset that includes at least two relevant clinical attributes. The method of any one of claims 1 to 6, further comprising the step of verifying against

8. A viral ARI classifier in Table 1, Table 2, Table 9, Table 10, and / or Table 12 Some of the genes listed (e.g., oligonucleotide probes homologous to said genes) 5, 10, 20, 30, or 50 to 80, 10 Any one of claims 1 to 7, comprising 0, 150 or up to 200 expression levels. A bacterial ARI classifier generated according to the method described in.

9. A viral ARI classifier in Table 1, Table 2, Table 9, Table 10, and / or Table 12 Some of the genes listed (e.g., oligonucleotide probes homologous to said genes) 5, 10, 20, 30, or 50 to 80, 10 Any one of claims 1 to 7, comprising 0, 150 or up to 200 expression levels. A viral ARI classifier generated according to the method described in.

10. A portion of the non-infectious disease classifiers in Table 1, Table 2, Table 9, Table 10, and / or Table 12 Genes listed as above (e.g., oligonucleotide probes homologous to the genes, 5, 10, 20, 30, or 50 to 80, 100 of the 150, 160, 170, 180, 200, 250, 300, 350, 400, 450, 500, 550, 600, 700, 800, 900, 1000, 150, 200, 350, 400, 550, 600, 700, 800, A non-communicable disease classifier constructed according to the methods described.

11. The etiology of the acute respiratory illness is selected from bacterial, viral, and / or non-infectious, A method for determining in a subject suffering from or at risk for the same, 、 (a) obtaining a biological sample from said subject; (b) determining the presence or absence of a gene sequence of a defined set of genes (i.e., a signature) in said biological sample; Measuring gene expression levels on the platform. (c) normalizing the gene expression levels to generate normalized gene expression values. (d) Bacterial acute respiratory infection (ARI) classifiers, viral ARI classifiers, and non-susceptible ARI classifiers The normalization genes are then normalized to one or more acute respiratory disease classifiers selected from a infectious disease classifier. inputting expression values, the classifier determining whether the classifier is A predefined weighting value (i.e., coefficient) for each gene in the predefined gene set. Optionally, the classifiers are retrieved from one or more databases. Tep, and (e) classifying bacterial ARI, viral ARI, or viral ARI based on the normalized gene expression values ​​and the classifier. Calculating the RI and the etiology probability for one or more of the non-communicable diseases. whereby the acute respiratory illness in the subject is determined to be bacterial or viral in origin. , non-infectious origin, or some combination thereof.

12. (f) determining whether the probability is greater than or equal to a predefined threshold, cutoff value, or range of values ​​indicative of likelihood of infection; and comparing the results with a range (e.g., a confidence interval). The method of claim 11 further comprising:

13. 13. The method of claim 11 or claim 12, wherein the subject is suffering from an acute respiratory illness. 。

14. 12. The method of claim 1, wherein the subject is suspected of having a bacterial or viral infection.

14. The method according to any one of 1 to 13.

15. If the sample does not indicate a probable bacterial ARI, a viral classifier and / or Repeating steps (d) and (e) using only the non-infected classifier to determine the presence or absence of Steps to determine whether an acute respiratory illness is of viral or non-infectious origin, or a combination The method of any one of claims 11 to 14, further comprising:

16. If the sample does not indicate a probable viral ARI, a bacterial classifier and / or Repeating steps (d) and (e) using only the non-infected classifier to determine the presence or absence of determining whether the acute respiratory illness is of bacterial origin, non-infectious origin, or a combination thereof; The method of any one of claims 11 to 14, further comprising:

17. If the sample does not indicate a probable non-infectious disease, the bacterial classifier and / or viral repeating steps (d) and (e) using only the sigma classifier to determine acute myelopathy in said subject; determining whether the respiratory illness is bacterial in origin, viral in origin, or a combination thereof. The method according to any one of claims 11 to 14, further comprising:

18. generating a report assigning a score to said subject indicative of a probability of said etiology of said acute respiratory illness; The method according to any one of claims 11 to 17, further comprising the step of:

19. The defined gene set comprises between 30 and 200 genes.

19. The method according to any one of claims 18 to 18.

20. The defined gene set is selected from the group consisting of those listed in Table 1, Table 2, Table 9, Table 10, and / or Table 12.

20. A method according to claim 11, comprising the steps of: The method described.

21. The biological sample is peripheral blood, sputum, nasopharyngeal swab, nasopharyngeal wash, bronchoalveolar lavage fluid, 21. Any of claims 11 to 20, selected from the group consisting of endotracheal aspirate, and combinations thereof. The method according to any one of claims 1 to 5.

22. The method according to any one of claims 11 to 20, wherein the biological sample is a peripheral blood sample. 。

23. The bacterial acute respiratory infection (ARI) classifier, viral ARI classifier, and non-immune ARI classifier The infectious disease classifier is obtained by the method according to any one of claims 1 to 7.

23. The method according to any one of claims 11 to 22.

24. A method of treating an acute respiratory disease in a subject, comprising administering to a subject a composition comprising any one of claims 11 to 23. administering to said subject an appropriate treatment regimen based on the etiology determined by the method of claim 1. A method including

25. If the etiology is determined to include a bacterial ARI, the appropriate treatment regimen includes an antibacterial 25. The method of claim 24, comprising administering

26. If the etiology is determined to include a viral ARI, the appropriate treatment regimen includes an antiviral 25. The method of claim 24, comprising virus treatment.

27. Have an acute respiratory illness selected from bacterial, viral, and / or non-infectious causes To monitor the response to a vaccine or drug in subjects who have or are at risk for determining a host response in said subject, said determining step comprising: A method carried out by the method according to any one of claims 11 to 23.

28. 28. The method of claim 27, wherein the drug is an antibacterial or antiviral drug.

29. Acute respiratory infection in subjects selected from bacterial, viral, and / or non-infectious 1. A system for determining the etiology of a disease, comprising: at least one processor; a sample input circuit configured to receive a biological sample from the subject; a processor coupled to said at least one processor and adapted to measure gene expression levels in said biological sample; a sample analysis circuit configured to determine input / output circuitry coupled to the at least one processor; A processor coupled to the at least one processor and configured to receive data, parameters, and / or a memory circuit configured to store the classifier; and When coupled to the processor and executed by the at least one processor, A computer embodied in a memory for causing the at least one processor to perform operations. A memory containing readable program code and wherein the operation comprises: Gene expression of a defined set of genes (i.e., a signature) in the biological sample. controlling / performing a current level measurement via said sample analysis circuit; normalizing the gene expression levels to generate normalized gene expression values; A predefined weighting value (i.e., Bacterial acute respiratory infection (ARI) classifier, viral ARI classifier, including coefficients and retrieving a non-infectious disease classifier from said storage circuit; The bacterial acute respiratory infection (ARI) classifier, the viral ARI classifier, and The normalization is performed to one or more acute respiratory disease classifiers selected from the non-infectious disease classifiers. Inputting gene expression values; Based on the classifier, one or more of bacterial ARI, viral ARI, and non-infectious illnesses are selected. Calculating the etiological probability for one or more of the following: whether the acute respiratory illness in the subject is of bacterial origin, viral origin, non-infectious origin, or or some combination thereof via said input / output circuit. thing Including, the system.

30. The system uses quantitative or semi-quantitative detection of gene expression to correlate with the etiology of the ARI.

30. The method of claim 29, further comprising computer readable code for converting the product score or probability to a system.

31. The system may include an array platform, a thermal cycler platform (e.g., (e.g., multiplexed and / or real-time PCR platforms), hybridization and multi-signal coded (e.g., fluorescent) detector platforms, nucleic acid mass spectrometry an analytical platform, a nucleic acid sequencing platform, or a combination thereof 31. The system of claim 29 or claim 30, comprising:

32. 29 to 30, wherein the defined gene set comprises between 30 and 200 genes.

31. A system as described in any one of claims 31.

33. The defined gene set is selected from the group consisting of those listed in Table 1, Table 2, Table 9, Table 10, and / or Table 12.

32. A method according to claim 29, comprising the steps of: The system described.

Images