Methods for diagnosing and treating acute respiratory infections

Abstract

To provide a method for determining etiology of acute respiratory illness selected from bacterial, viral, and / or non-infectious.SOLUTION: A method comprises: (a) obtaining a biological sample from a subject; (b) measuring on a platform, gene expression levels of a pre-defined set of genes in the biological sample; (c) normalizing the gene expression levels to generate normalized gene expression values; (d) entering the normalized gene expression values into one or more acute respiratory illness classifiers selected from a bacterial acute respiratory infection (ARI) classifier, a viral ARI classifier and a non-infectious illness classifier, the classifier comprising pre-defined weighting values for each of the genes of the pre-defined set of genes for the platform; and (e) calculating an etiology probability for one or more of a bacterial ARI, viral ARI and non-infectious illness on the basis of the normalized gene expression values and the classifier.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] Related applications This application is U.S. Provisional Patent Application No. 62 / 187,683, filed on July 1, 2015. U.S. Provisional Patent Application No. 62 / 257,406, filed on November 19, 2015. Claiming the benefits of and disclosing those U.S. provisional patent applications, by reference, is part of this specification. It shall be.

[0002] Explanation of federal government financial support This invention is from the National Institutes of Health (NIH) Federal grant numbers U01AI066569, P20RR016480, awarded by ) and HHSN266200400064C, and Defense Advance Awarded by the Research Projects Agency (DARPA) Federal grant numbers N66001-07-C-2024 and N66001-09-C-2 This invention was made with government assistance under 082. The U.S. government has certain rights to this invention. [Background technology]

[0003] Acute respiratory infections are common in emergency medical settings and have a significant mortality and morbidity rate worldwide. , and economic losses. Respiratory tract infections or acute respiratory infections (ARIs) are 2 In 2011, it caused 3.2 million deaths worldwide, more than any other cause. , and 164 million disability-adjusted life years were lost (World Health Organization) (rganization., 2013a, 2013b). In 2012, the world The fourth leading cause of death is lower respiratory tract infections, which are less accessible to low- and middle-income individuals who have limited access to supportive care. In the country, lower respiratory tract infections are the leading cause of death (WHO factsheet, accessed August 22, 2014). These diseases are also a problem in developed countries. In the United States in 2010, the Centers for Disease Control (CDC) determined that pneumonia and influenza alone resulted in 15.1 deaths per 100,000 people in the US population. The elderly and children under 5 years old are particularly prone to poor prognosis due to ARI. For example, in 2010, pneumonia accounted for 18.3% of all deaths worldwide in children under 5 years old (i.e., almost 1.4 million deaths). These diseases are also a problem in developed countries. In the United States in 2010, the Centers for Disease Control (CDC) determined that pneumonia and influenza alone resulted in 15.1 deaths per 100,000 people in the US population. The elderly and children under 5 years old are particularly prone to poor prognosis due to ARI. For example, in 2010, pneumonia accounted for 18.3% of all deaths worldwide in children under 5 years old (i.e., almost 1.4 million deaths). In the United States in 2010, the Centers for Disease Control (CDC) determined that pneumonia and influenza alone resulted in 15.1 deaths per 100,000 people in the US population. The elderly and children under 5 years old are particularly prone to poor prognosis due to ARI. For example, in 2010, pneumonia accounted for 18.3% of all deaths worldwide in children under 5 years old (i.e., almost 1.4 million deaths). In 2010, the Centers for Disease Control (CDC) determined that pneumonia and influenza alone resulted in 15.1 deaths per 100,000 people in the US population. The elderly and children under 5 years old are particularly prone to poor prognosis due to ARI. For example, in 2010, pneumonia accounted for 18.3% of all deaths worldwide in children under 5 years old (i.e., almost 1.4 million deaths). The elderly and children under 5 years old are particularly prone to poor prognosis due to ARI. For example, in 2010, pneumonia accounted for 18.3% of all deaths worldwide in children under 5 years old (i.e., almost 1.4 million deaths). The elderly and children under 5 years old are particularly prone to poor prognosis due to ARI. For example, in 2010, pneumonia accounted for 18.3% of all deaths worldwide in children under 5 years old (i.e., almost 1.4 million deaths). For example, in 2010, pneumonia accounted for 18.3% of all deaths worldwide in children under 5 years old (i.e., almost 1.4 million deaths). For example, in 2010, pneumonia accounted for 18.3% of all deaths worldwide in children under 5 years old (i.e., almost 1.4 million deaths).

[0004] Pneumonia and other lower respiratory tract infections can be mainly due to many different pathogens that are mainly viruses, bacteria, or less frequently fungi. Among the viral pathogens, influenza is one of the most notorious due to the constant concern about new strains (e.g., avian influenza) that cause a different number of affected individuals, different severities depending on the season, and much higher morbidity and mortality. However, among the viral pathogens, influenza is just one of several that cause significant human diseases. Respiratory syncytial virus (RSV) is the leading cause of hospitalization of children in developed countries during the winter. In children under 5 years old in 2005, about 33 million new cases of RSV infection were reported worldwide, and 3.4 million were severe enough to be hospitalized. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. Pneumonia and other lower respiratory tract infections can be mainly due to many different pathogens that are mainly viruses, bacteria, or less frequently fungi. Among the viral pathogens, influenza is one of the most notorious due to the constant concern about new strains (e.g., avian influenza) that cause a different number of affected individuals, different severities depending on the season, and much higher morbidity and mortality. However, among the viral pathogens, influenza is just one of several that cause significant human diseases. Respiratory syncytial virus (RSV) is the leading cause of hospitalization of children in developed countries during the winter. In children under 5 years old in 2005, about 33 million new cases of RSV infection were reported worldwide, and 3.4 million were severe enough to be hospitalized. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. Among the viral pathogens, influenza is one of the most notorious due to the constant concern about new strains (e.g., avian influenza) that cause a different number of affected individuals, different severities depending on the season, and much higher morbidity and mortality. However, among the viral pathogens, influenza is just one of several that cause significant human diseases. Respiratory syncytial virus (RSV) is the leading cause of hospitalization of children in developed countries during the winter. In children under 5 years old in 2005, about 33 million new cases of RSV infection were reported worldwide, and 3.4 million were severe enough to be hospitalized. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. Among the viral pathogens, influenza is one of the most notorious due to the constant concern about new strains (e.g., avian influenza) that cause a different number of affected individuals, different severities depending on the season, and much higher morbidity and mortality. However, among the viral pathogens, influenza is just one of several that cause significant human diseases. Respiratory syncytial virus (RSV) is the leading cause of hospitalization of children in developed countries during the winter. In children under 5 years old in 2005, about 33 million new cases of RSV infection were reported worldwide, and 3.4 million were severe enough to be hospitalized. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. Among the viral pathogens, influenza is one of the most notorious due to the constant concern about new strains (e.g., avian influenza) that cause a different number of affected individuals, different severities depending on the season, and much higher morbidity and mortality. However, among the viral pathogens, influenza is just one of several that cause significant human diseases. Respiratory syncytial virus (RSV) is the leading cause of hospitalization of children in developed countries during the winter. In children under 5 years old in 2005, about 33 million new cases of RSV infection were reported worldwide, and 3.4 million were severe enough to be hospitalized. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. However, among the viral pathogens, influenza is just one of several that cause significant human diseases. Respiratory syncytial virus (RSV) is the leading cause of hospitalization of children in developed countries during the winter. In children under 5 years old in 2005, about 33 million new cases of RSV infection were reported worldwide, and 3.4 million were severe enough to be hospitalized. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. However, among the viral pathogens, influenza is just one of several that cause significant human diseases. Respiratory syncytial virus (RSV) is the leading cause of hospitalization of children in developed countries during the winter. In children under 5 years old in 2005, about 33 million new cases of RSV infection were reported worldwide, and 3.4 million were severe enough to be hospitalized. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. Respiratory syncytial virus (RSV) is the leading cause of hospitalization of children in developed countries during the winter. In children under 5 years old in 2005, about 33 million new cases of RSV infection were reported worldwide, and 3.4 million were severe enough to be hospitalized. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. In children under 5 years old in 2005, about 33 million new cases of RSV infection were reported worldwide, and 3.4 million were severe enough to be hospitalized. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. This viral infection alone is estimated to cause 66,000 - 199,000 children to lose their lives every year. And in the United States alone, in the population over 65 years old, about 10,000 cases occur every year. The deaths are linked to RSV infection. In addition to known viral pathogens, novel and emerging pathogens are also linked. History shows that infectious diseases can appear at any time and spread globally within days or weeks. Recent examples include the SARS coronavirus, which was prevalent from 2003 to 2000. When it first appeared in 2004, it had a 10% mortality rate. More recently, Middle East Respiratory Syndrome (MER) has emerged. S) The coronavirus continues to be on the verge of exploding in the Middle East, and deaths are increasing. The rate is 30%. Both of these infections present with respiratory symptoms, and initially, It may be indistinguishable from other ARIs.

[0005] Viral infections cause the majority of ARIs, but bacterial etiologies also play a role, particularly in lower respiratory tract infections. This is evident in the context of the disease. The specific causes of bacterial ARI are geographically and clinically. It varies depending on the situation, but these include Streptococcus pneumoniae (pN). eumoniae, Staphylococcus aureus Haemophilus influenzae, Chlamydia pneumoniae Chlamydia pneumoniae, Mycoplasma pneumonia asma pneumoniae), Klebsiella pneumoniae (iae), Escherichia coli, and Pseudomonas aeruginosa This includes Monas aeruginosa. Identifying these pathogens in culture is important. It relies on their proliferation, which typically takes several days, and there is no sense of detection of infectious material. There are limitations to this. It is difficult to obtain enough samples for testing: 1 with community-acquired pneumonia In a study of 669 patients, only 14% produced "good" positive cultures. We were able to provide sputum samples (Garcia-Vazquez et al., 20 04). Clinicians are anxious because they are aware of the limitations of these trials, and as a result... Therefore, antibacterial treatment is frequently prescribed without any confirmation of a bacterial infection.

[0006] The ability to rapidly diagnose the pathogenesis of ARIs allows for optimal treatment of individual patients. Epidemiological surveillance to identify and track outbreaks and epidemics, and the rise in antimicrobial resistance This involves widespread social importance at multiple levels, including guidance on the appropriate use of antimicrobial agents to stop disease. This is a serious, urgent global problem. Early and appropriate antimicrobial treatment is crucial for patients with severe infections. It is well established that this improves the prognosis in patients. This is partly due to excessive antibiotic treatment. To encourage use. 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, However, it is estimated that only a small number of these patients require antibacterial treatment (Ca ntrell et al. 2003, Clin. Ther. Jan;24(1 ):170-82). A similar trend is observed in the emergency department. Presence of viral pathogens. Even if it has been confirmed microbiologically, this does not rule out the possibility of a simultaneous bacterial infection. As a result, antibacterial agents are often prescribed "just in case." This spiral Empiricism, which falls into the trap of ignorance, contributes to the rise of antimicrobial resistance (Gould, 2009; Kim (Gallis, 1989), itself, higher mortality rate, longer hospital stay, and health This involves the cost of scare (Cosgrove 2006, Clin. Infect. D is., Jan 15;42 Suppl 2:S82-9). In addition, antibiotics Proper use is important to avoid adverse drug reactions and other complications, such as Clostridium erythrorhizon. It can cause diarrhea associated with Clostridium difficile. (Zaas et al., 2014).

[0007] Acute respiratory infections are associated with many different diseases, including those not caused by infection. It is often characterized by persistent nonspecific symptoms (e.g., fever or cough). Existing diagnostic methods for radioisotopes (RI) are insufficient in many respects. Conventional microbiological tests are limited. Due to low sensitivity and specificity, slow turnaround time, or the complexity of the test Restricted (Zaas et al. 2014, Trends Mol Med) 20(10):579-88). Current tests for detecting specific viral pathogens, for example One limitation of assays based on multiplex PCR is the inability to detect emergency or generalized virus strains. It is impossible. An influenza pandemic is a new type of pandemic in which a population has no natural resistance. It occurs when a virus spreads. Influenza pandemics can be devastating. Many. For example, during the Spanish flu pandemic of 1918-1919, approximately 20-40% of the world's population was affected. It affected % and killed approximately 50 million people; the Asian flu in 1957-1958 affected approximately 200 Millions of people died; the Hong Kong flu of 1968-1969 killed approximately 1 million people; Then, in 2009-2010, approximately 43 to 89 million people contracted swine flu. The disease has caused 8,870 to 18,300 related deaths, according to Centers for Disease Control and Prevention. Isease Control estimates that the emergence of these new strains is due to their This poses a challenge for existing diagnostic methods that are not designed to detect infection. Approval is required over several days and at a specialized testing center such as the State Department of Health or the CDC. This was particularly evident during the 2009 influenza pandemic, which was only conducted in [location]. (Kumar & Henrickson 2012, Clin Microb) iol Rev 25(2):344-61). Ebola virus disease in West Africa. The current pandemic presents similar challenges. Furthermore, the future of infectious diseases Since the pandemic is unavoidable, it is highly likely that we will continue to face this problem.

[0008] Further limitations of diagnostic methods using a testing paradigm for specific viruses or bacteria Even if pathogenic microorganisms can be detected, this is because the patient's symptoms are not detected. This does not prove that it is caused by a pathogen. Microorganisms form colonies. It may be present as part of the normal bacterial flora of an individual, or it has been tested. It can be detected by contamination of the sample (e.g., nasal swab or lavage solution). Viruses and cells Recently approved multiplex PCR assays, including those for detecting bacteria, offer high sensitivity, however These tests do not distinguish between asymptomatic carriers of the virus and true infection. For example, ARI odor Furthermore, 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 one pathogen is detected, the disease may not be available or implemented. It may be due to a second pathogen for which no tests were available.

[0009] The report indicates that the host gene expression profile that distinguishes viral ARI from healthy controls is (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 or bacterial infections affecting health. Distinguishing between bacterial and viral ARIs is a more clinically meaningful distinction than detection. No separate procedure is performed (Hu, Yu, Crosby, & Storch, 2013; P Arnell et al., 2012; Ramilo et al., 2007 ).

[0010] Thus, current diagnostic methods include bacterial infections, viral infections, and non-infectious causes. The ability to distinguish between symptoms arising from, or the ability to identify co-infections of bacteria and viruses. There are limits to it. [Overview of the Initiative]

[0011] This disclosure, in part, overcomes many of the limitations of current methods for determining the etiology of respiratory symptoms. We offer molecular diagnostic tests that analyze patterns of co-expressed genes or signatures. By measuring and analyzing these factors, the host's response to infectious agents is detected. The gene expression signature is found in humans or animals exhibiting symptoms consistent with acute respiratory infections. Or (for example, during an epidemic or localized disease outbreak) squirrels that develop acute respiratory infections It can be measured in blood samples from humans or animals that have a prodromal state (e.g., prodromal symptoms). The measurement of host responses as taught herein is for bacterial ARI, viral ARI. I distinguish between non-infectious etiologies and ARIs resulting from co-infection with bacteria and viruses. It is possible to release it.

[0012] This multi-component study functions with unprecedented precision and clinical applicability, and is beneficial to healthcare providers. This uses the host's (subject's or patient's) response to determine the properties of the infectious agent at the pathogen class level. In individual patients who show symptoms that are not specific to those factors alone, the condition is determined to be definitively determined by the diagnosis. It makes it possible to eliminate the infectious cause of the condition. In some embodiments, the result is , does not detect respiratory viruses or bacterial species (i.e., does not distinguish between viruses and bacteria) (This distinguishes between specific genera or species of viruses or bacteria.) It contains probes or reagents for pathogens, and therefore detects only those specific pathogens. It offers advantages over the current examination, which is limited to issuing results.

[0013] One aspect of this disclosure is whether the acute respiratory symptoms in the subject are of bacterial or viral origin, A method for determining whether the source is non-infectious or (a) obtaining a biological sample from the subject. Step (b) Obtain the gene expression profile of the target from the biological sample, which is called the signature. The step of determining by evaluating the expression levels of a defined set of genes, (c Regarding the technology (i.e., platform) used to perform the aforementioned measurements: (d) The steps to normalize the gene expression level and generate a normalized value for each signature A bacterial taxonomicon having a predefined weighting value (coefficient) for each of its genes, The step of inputting the normalized values ​​into the virus classifier and / or non-infectious disease classifier, (e) The output of the classifier is defined as a threshold, cutoff value, or value indicating the likelihood of infection. (f) using the output, the patient who provided the sample Having an infection of bacterial or viral origin, or a non-communicable disease or condition thereof. The step includes determining whether it has several combinations of them, consisting of them or not. It provides a way to become essentially.

[0014] Another aspect of this disclosure is that acute respiratory infections (ARIs) in the subject are of bacterial origin or are viral. A method for determining whether the origin is Rus or non-infectious, wherein (a) the organism (b) Obtain a biological sample, and determine the gene expression profile of the target gene from the biological sample. (c) The step of determining by evaluating the expression level of the imputed gene set, Regarding the technology (i.e., platform) used to perform the measurement, genetics (d) Normalizing the child expression level to obtain a normalized value, and the legacy in each signature The normalized values ​​are input to a classifier that has predefined weighted values ​​for each of the genes. (e) The output of the classifier is defined as a threshold, cutoff value, or probability of infection (f) If the sample is negative for bacteria, This involves repeating step (d) using only virus classifiers and non-infectious classifiers. (g) a method for classifying the sample as either a viral or non-infectious disease. It provides a method that includes steps, consists of them, or is essentially made from them.

[0015] Another aspect of this disclosure is that acute respiratory infections (ARIs) in the subject are of bacterial origin or are viral. A method for determining whether the origin is Rus or non-infectious, wherein (a) the organism (b) Obtain a biological sample, and determine the gene expression profile of the target gene from the biological sample. (c) The step of determining by evaluating the expression level of the imputed gene set, Regarding the technology (i.e., platform) used to perform the measurement, genetics (d) Normalizing the child expression level to obtain a normalized value, and the legacy in each signature The normalized values ​​are input to a classifier that has predefined weighted values ​​for each of the genes. (e) The output of the classifier is defined as a threshold, cutoff value, or probability of infection (f) If the sample is negative for the virus, In the case of a sample, repeat step (d) using only bacterial and non-infectious classifiers. (g) Steps for classifying the sample as a bacterial pathogen or a non-infectious disease. To provide a method that includes, consists of, or is essentially composed of.

[0016] Another aspect of this disclosure is that acute respiratory infections (ARIs) in the subject are of bacterial origin or are viral. A method for determining whether the origin is Rus or non-infectious, wherein (a) the organism (b) Obtain a biological sample, and determine the gene expression profile of the target gene from the biological sample. (c) The step of determining by evaluating the expression level of the imputed gene set, Regarding the technology (i.e., platform) used to perform the measurement, genetics (d) Normalizing the child expression level to obtain a normalized value, and the legacy in each signature The normalized values ​​are input to a classifier that has predefined weighted values ​​for each of the genes. (e) The output of the classifier is defined as a threshold, cutoff value, or probability of infection (f) The step of comparing the value to a range of values ​​indicating the absence of infection. If so, repeat step (d) using only the virus classifier and bacterial classifier. (g) the steps of (g) classifying the sample as either viral or bacterial in origin. It provides a way of including, consisting of, or essentially being composed of, steps.

[0017] Another aspect of this disclosure relates to acute respiratory infections (ARIs) of unknown etiology in the subject. A method for treating ) the subject, comprising the steps of (a) obtaining a biological sample from the subject, and (b) biological From the target sample, the gene expression profile of the target gene is obtained from a defined set of genes (for example, one, This is determined by evaluating the expression levels of two, three, or more signatures. (c) the step of determining the measurement (i.e., the technology used to perform the measurement) The step of normalizing the gene expression level for the set form to obtain a normalized value, d) Details with a defined weighted value for each gene in each signature. Steps to input the normalized values ​​into the bacterial classifier, virus classifier, and non-infectious disease classifier. (e) The output of the classifier indicates a predefined threshold, cutoff value, or likelihood of infection. (f) comparing the sample with a range of values, (f) determining whether the sample is a bacterial pathogen, a viral pathogen, or (g) The step of classifying it as a non-infectious disease, and (e) the steps identified by step (e) This includes the step of administering an appropriate treatment regimen to the subject as described above, and consists of the following: Or provide a way to essentially derive from them. In some embodiments, step (g) If it is determined that the etiology of ARI is bacterial, this includes administering antibacterial treatment. In another embodiment, step (g) is determined that the etiology of ARI is viral. This may include administering antiviral treatment.

[0018] Another aspect is acute respiratory illness selected from bacterial, viral, and / or non-infectious diseases. Response to vaccines or drugs in individuals who have or are at risk of developing a disease. A method for monitoring the host response of the subject, comprising the step of determining the host response of the subject, The steps are carried out in the manner taught herein. In one embodiment, the drug is an antibacterial or antiviral agent.

[0019] In some embodiments of the above-described model, the method shows the probability of the etiology of ARI. This further includes the step of creating a report that assigns A to the target.

[0020] Acute respiratory illness in subjects selected from bacterial, viral, and / or non-infectious conditions. A system for determining the etiology of a disease, comprising at least one processor; derived from the subject. A sample input circuit configured to receive a biological sample; the at least one processor A sample connected to and configured to determine the gene expression level of the biological sample. Analysis circuit; input / output circuit connected to at least one processor; at least It is also linked to a single processor and stores data, parameters, and / or classifiers. A memory circuit configured to do the following; and connected to the processor, and the least When both are executed by one processor, the operation is performed on at least one of the processors. To cause, among the memory containing computer-readable program code that is implemented in memory This includes one or more of the following (including combinations thereof), wherein the operation is performed on the biological sample. The measurement of gene expression levels of a predefined set of genes (i.e., signatures) is performed before Control / implement via the sample analysis circuit; normalize the gene expression level, To produce a gene expression level; for each gene in a defined set of genes A classifier for bacterial acute respiratory infections (ARI) that includes predefined weighted values ​​(i.e., coefficients). Retrieving viral ARI classifiers and non-infectious disease classifiers from the memory circuit. ; Bacterial acute respiratory infection (ARI) classifiers, viral ARI classifiers, and non-infectious The normalized gene expression is applied to one or more acute respiratory disease classifiers selected from the disease classifiers. Enter a value; based on the classifier, bacterial ARI, viral ARI, and non To calculate the probability of etiology for one or more infectious diseases; and to apply the above to In this case, is the acute respiratory illness of bacterial origin, viral origin, non-infectious origin, or a combination of these? The system includes controlling the output of the decision of one of several combinations via the input / output circuit. More stems will be provided.

[0021] In some embodiments, the system enables quantitative or semi-quantitative detection of gene expression. Includes computer-readable code that converts ARIs into cumulative scores or probabilities of etiology.

[0022] In some embodiments, the system includes an array platform, a thermal cycle Ra platform (e.g., multiplexing and / or real-time PCR platform) (Violet), hybridization, and multi-signal encoding (e.g., fluorescence) detectors Platform, nucleic acid mass spectrometry platform, nucleic acid sequencing platform This includes mu, or combinations thereof.

[0023] In some embodiments of the above-described model, the defined gene set is at least three It includes the gene signature.

[0024] In some embodiments of the above-described model, the biological sample is peripheral blood, sputum, or nasopharyngeal swab. , nasopharyngeal lavage solution, bronchoalveolar lavage solution, tracheal aspirate, and a group consisting of combinations thereof Includes selected samples.

[0025] In some embodiments of the above-described model, the bacterial classifiers are as shown in Tables 1, 2, 9, and 10. and / or genes listed as part of the bacterial taxonomy in Table 12 (e.g., the aforementioned (Measurable by oligonucleotide probes homologous to a gene or gene transcript) 5, 10, 20, 30, or 50 to 80, 100, 150, and This includes expression levels up to 200. In some embodiments, the virus classifier is Tables 1, 2, 9, 10, and / or 12 as part of the virus classifier The listed genes (for example, oligonucleotides homologous to the gene or gene transcript) With a probe, you can measure 5, 10, 20, 30, or 50 of the measurable points, up to 8. Includes expression levels of 0, 100, 150, or up to 200. Several implementations In this context, the non-infectious disease classifiers are found in Tables 1, 2, 9, 10, and / or 12. Genes listed as part of the non-infectious disease classifiers in (for example, the aforementioned genes or Five of the oligonucleotide probes homologous to gene transcripts (measurable), 10 From 1, 20, 30, or 50 to 80, 100, 150, or 200 This includes the expression level in [location].

[0026] (a) Means for extracting mRNA from biological samples; (b) Tables 1, 2, 9, and 9 10, and / or 8 from 5, 10, 20, 30, or 50 from Table 12 Regions homologous to transcripts from 0, 100, 150, or 200 genes For preparing one or more arrays consisting of multiple synthetic oligonucleotides (c) means; and (c) instructions for use, including, consisting of, or essentially consisting of Furthermore, a kit for determining the etiology of acute respiratory infections (ARIs) in the target population is also provided. It will be done.

[0027] Another aspect of this disclosure involves using a kit to evaluate acute respiratory infection (ARI) classifiers. A method using (a) 5 from Table 1, Table 2, Table 9, Table 10, and / or Table 12 10, 10, 20, 30, or 50 to 80, 100, 150, or 2 A single compound consisting of multiple synthetic oligonucleotides having homologous regions to up to 00 genes. (b) A step of creating multiple arrays; (b) an oligo having a region homologous to the normalized gene. (c) Adding nucleotides to the array; (c) Suffering from an acute respiratory infection (ARI) (d) A step of obtaining a biological sample from the subject; (d) Isolating RNA from the sample and Steps to construct a ranscriptome; (e) (for example, proportional to the level of gene expression) (By measuring fluorescence or current, etc.) the transcriptome on the array (f) the steps of measuring the transcriptome measurement against the normalized gene The normalized measurements are then electronically transferred to a computer for normalization and classification. (g) the step of preparing a report, and optionally, (h) appropriate based on the results said above. Provide a method that includes, consists of, or is essentially derived from, the steps of performing a procedure. do.

[0028] In some embodiments, the method includes at least two related clinical attributes. Further steps include externally validating the ARI classifier against knowledge datasets. In this embodiment, the dataset includes GSE6269, GSE42026, and GSE40 Select from the group consisting of 396, GSE20346, GSE42834, and combinations thereof. It will be selected.

[0029] Further aspects of this disclosure provide all that is disclosed and illustrated herein.

[0030] Regarding the treatment of acute respiratory infections (ARIs) in patients with unknown etiologies, The use of the ARI classifier as instructed in the specification is also provided.

[0031] The aforementioned aspects and other features of this disclosure are shown herein in reference to the accompanying drawings. This is explained in the following description. [Brief explanation of the drawing]

[0032] [Figure 1] This is a schematic diagram illustrating a method (training 10) for obtaining classifiers according to some embodiments of the present disclosure, where each classifier consists of a weighted sum of all or a subset of normalized gene expression levels. This weighted sum defines the probability that enables a decision (classification), in particular, compared to a threshold or confidence interval. The exact combination of genes, their weights, and thresholds for each classifier obtained through training are specific to a particular platform. The classifiers (or more precisely, their components, i.e., weights and thresholds or confidence intervals (values)) form a database. Weights with non-zero values ​​determine the subset of genes used by the classifier. The process is repeated to obtain all three classifiers (bacterial ARI, viral ARI, and non-infectious ARI) within a specified platform for matching gene expression values. [Figure 2] This figure shows examples of how to create and / or use classifiers according to some embodiments of the present disclosure. [Figure 3] This is a schematic diagram illustrating a method for classifying the etiology of acute respiratory symptoms suffered by a subject, using a classifier according to some embodiments of the present disclosure. [Figure 4] This is a schematic diagram illustrating decision patterns for using secondary classification to determine the pathogenesis of ARI in a subject, according to some embodiments of the present disclosure. [Figure 5] This is a diagram illustrating the training method example presented in Example 1. Classifiers for each condition were developed using a cohort of patients encompassing bacterial ARI, viral ARI, or non-infectious diseases. This integrated ARI classifier was validated using one-out cross-validation and compared to three published classifiers for bacterial versus viral infection. The integrated ARI classifier was also externally validated on six publicly available datasets. In one experiment, healthy volunteers were included in the training set and their suitability was determined as "non-infectious" controls. All subsequent experiments were conducted without this healthy control cohort. [Figure 6]This graph shows the results of a one-case cross-validation of three classifiers (bacterial ARI, viral ARI, and non-infectious disease) according to the training method example presented in Example 1. Each patient is assigned a probability of having bacterial ARI (triangle), viral ARI (circle), or non-infectious disease (square). Patients clinically determined to have bacterial ARI, viral ARI, or non-infectious disease are presented in the upper, middle, and lower panels, respectively. The overall classification accuracy was 87%. [Figure 7] This graph shows the evaluation of healthy adults as non-infected controls rather than as controls with the disease but not infection. This figure demonstrates the unexpected superiority of using diseased but non-infected controls. [Figure 8] This figure shows the positive and negative predictive values ​​for A) bacterial ARI classification and B) viral ARI classification as functions of prevalence. [Figure 9] This Venn diagram shows the overlaps in bacterial ARI classifiers, viral ARI classifiers, and non-infectious disease classifiers. There are 71 genes in the bacterial ARI classifier, 33 genes in the viral ARI classifier, and 26 genes in the non-infectious disease classifier. One gene overlaps between the bacterial and viral ARI classifiers. Five genes overlap between the bacterial and non-infectious disease classifiers. Four genes overlap between the viral and non-infectious disease classifiers. [Figure 10]This graph shows the classifier performance in patients with co-infections, identified by bacterial and viral pathogens. Bacterial ARI classifiers and viral ARI classifiers were trained on subjects (GSE60244) with either bacterial infection (n=22) or viral infection (n=71). This same dataset also included 25 subjects with bacterial / viral co-infections. As shown in the figure, bacterial and viral classifier predictions were normalized to the same scale. Each subject receives two probabilities: the probability of a bacterial ARI host response and the probability of a 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, potentially indicating 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]This is a block diagram of a classification system and / or computer program product that may be used on a platform. The classification system and / or computer program product 1100 may include a processor subsystem 1140, which includes 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, which may be either electrically interconnected or disconnected. The processor 1140 is configured to execute computer program code from a memory device, such as memory 1150, to perform at least some of the operations and methods described herein. A storage circuit 1170 may store a database that provides access to data / parameters / classifiers used by the classification system 1100, such as signatures, weights, and thresholds. An input / output circuit 1160 may include a display and / or a user input device, such as a keyboard, touchscreen, and / or pointing device. Devices attached to the input / output circuit 1160 may be used by a user of the classification system 1100 to provide information to the processor 1140. The devices attached to the input / output circuit 1160 may include a network controller or communication controller, input devices (such as a keyboard, mouse, or touchscreen), and output devices (such as a 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 code executed by the processor 1140 and stored in memory 1150 and / or storage circuit 1170. Updates provided through the update circuit 1180 may also include updates to parts of the storage circuit 1170 related to a database and / or other data storage format that maintains information about the classification system 1100, such as signatures, weights, and thresholds. The sample input circuit 1110 provides an interface for the classification system 1100 to receive biological samples to be analyzed.The sample processing circuit 1120 may further process the biological sample within the classification system 1100 to prepare the biological sample for automated analysis. [Modes for carrying out the invention]

[0033] For the purpose of facilitating understanding of the principles of this disclosure, preferred embodiments are now referenced. A special language is used to describe it. However, the scope of this disclosure is limited to that. It is not intended to be, and such as the disclosure illustrated herein. It is understood that changes and further modifications are intended to be made in a manner that would ordinarily be conceived by those skilled in the art. It will probably happen.

[0034] The articles "a" and "an" refer to one of the grammatical objects of the article. In this specification, the term is used to refer to more than one (i.e., at least one). For example, "one element" means at least one element and does not include more than one element. obtain.

[0035] Unless otherwise specified, all technical terms used herein are common to those skilled in the art. It has the same meaning as being understood as such.

[0036] This disclosure relates to genes in the blood in response to exposure to pathogens that cause acute respiratory infections. Changes in protein and metabolite expression can be identified with a high degree of precision as the etiology of ARI in the subject. It provides that it can be used to define and characterize.

[0037] definition As used herein, the terms “acute respiratory infection” or “ARI” often Symptoms consistent with upper or lower respiratory tract infections caused by bacterial or viral pathogens, and / or physical findings (e.g., symptoms such as cough, wheezing, fever, sore throat, congestion; elevated heart rate) Symptoms include elevated respiratory rate, abnormal white blood cell count, and low arterial carbon dioxide partial pressure (PaCO2). It presents with physical findings and is characterized by a rapid progression of symptoms over several hours to several days. It refers to an infectious disease or illness. ARIs mainly affect the upper respiratory tract (URI) and lower respiratory tract (LRI). Or it could be a combination of the two. ARIs are used to control the spread of infection beyond the respiratory tract. Systemic effects may occur due to collateral damage induced by or by an immune response. Examples of the former include Staphylococcus aureus pneumonia that has spread into the bloodstream, and endocarditis (infection of the heart valves). It can cause secondary infections, including septic arthritis (joint infection), or osteomyelitis (bone infection). The latter example is influenza pneumonia that leads to acute respiratory distress syndrome and respiratory failure. include.

[0038] As used herein, the term "signature" means a specific combination of organisms that have been identified. A set of biological analytes representing the presence or absence of a biological state, and the measurement of said analytes. This refers to the possible amount. These signatures have a known state (e.g., confirmed respiration). Multiple pairs (having bacterial infections of the organs, viral infections of the respiratory tract, or suffering from non-communicable diseases) It has been found in elephants, and one or more categories or results of interest ( Identify them individually or collectively. These measurable markers are also known as biological markers. The analytes can be measured at the gene expression level, protein or peptide level, or metabolite level. It could be (but not limited to) a bell. Courchesne et al. U.S. Patent U.S. Patent Publication No. 2015 / 0227681; U.S. Patent Publication No. 2016 / 01539 by Eden et al. See also issue 93.

[0039] In some embodiments as disclosed herein, “Signature” is generated If the current level is incorporated into a classifier as taught herein, bacterial ARI A specific combination of genes that identifies a condition such as viral ARI or non-infectious disease. For example, see Tables 1, 2, 9, 10, and 12 below. In application, the signature does not detect respiratory viruses or bacterial species. In other words, the distinction between viruses and bacteria is made between specific genera of viruses or bacteria. (or does not distinguish between species), and / or does not perceive the specific causes of non-infectious diseases. .

[0040] As used herein, the terms “classifier” and “predictor” r) is used interchangeably and is intended to assign to a category based on predetermined observation results. To produce a score for the result or for individual patients, the signature value (for example, defined) The gene expression levels for each gene set and the determined values ​​for each signature component. This refers to a mathematical function that uses coefficients (or weights). The classifier is linear and / or stochastic. This is possible. The score is a function of the sum of signature values ​​weighted by a series of coefficients. In this case, the classifier is linear. Furthermore, if the function of the signature value generates a probability, The classifier is probabilistic, and its value is between 0 and 1.0 (or between 0% and 100%). Each subject or observation will belong to a specific category or produce a specific result. Quantify the probabilities. Probit regression analysis and logistic regression analysis generate probabilities. To achieve this, the probit function and the logistic link function are used, respectively. This is an example of a linear classifier.

[0041] Classifiers like those taught in this specification are obtained through a procedure known as "training". This can be done, and the procedure involves known category memberships (e.g., bacterial ARI, u A dataset containing observational results regarding viral ARIs and / or non-infectious diseases. Uses a specific signature. See Figure 1. Specifically, training involves, in addition to the optimal signature, a predetermined signature. The optimal coefficient (i.e., weight) for each component of the tea (e.g., gene expression level component) The goal is to find the optimal result, which is determined by the highest achievable classification accuracy. .

[0042] "Classification" is used to select one subject who has or is at risk of having acute respiratory symptoms. or multiple categories or outcomes (for example, the patient is infected with the pathogen or is not infected) Another categorization is that patients are infected with viruses and / or bacteria. This refers to a method of assigning to (which may be). See Figure 3. In some cases, the target is It may be classified into more than one category (for example, in the case of a simultaneous infection of bacteria and viruses). The result or category is determined by the score value provided by the classifier, The value can be compared to a cutoff value or threshold, a confidence level or limit. And the probability of belonging to a particular category can be shown (for example, where the classifier reports the probability). ).

[0043] As used herein, the term "indicates" when used in conjunction with gene expression levels. "Dichotomous" refers to a gene expression level that is in an alternative biological state (e.g., bacterial). Upregulation is also possible compared to the expression levels in ARI (or viral ARI) or the control. This means that it is downregulated, modified, or altered. Protein level The term "show" when used together with "show" means that the protein level is the standard protein level or It is higher or lower, increased or decreased, compared to the level in the alternative biological state. It means that it has been slightly modified or changed.

[0044] The terms “subject” and “patient” are used interchangeably when examining, studying, or treating. This refers to any animal that is subject to this disclosure. Not intended. In some embodiments of the present invention, humans are preferred subjects, but others In the embodiments, non-human animals are preferred subjects, including mice, monkeys, and k Roast weasel, cow, sheep, goat, pig, chicken, turkey, dog, cat, rooster This includes, but is not limited to, mackerel and reptiles. In a particular embodiment, The subjects are individuals who have ARI or are exhibiting ARI-like symptoms.

[0045] Where used herein, “Platform” or “Technology” means “Disclosure.” Accordingly, a signature, for example, an instrument that can be used to measure gene expression levels. For example, instruments and related parts, computers, and one or more as taught herein. This refers to computer-readable media containing multiple databases, reagents, etc. Examples of M include array platforms and thermal cycler platforms (for example, Multiplexing and / or real-time PCR platform), nucleic acid sequencing platform Forms, hybridization, and multi-signal encoding (e.g., fluorescence) Detector platforms, nucleic acid mass spectrometry platforms, magnetic resonance platforms This includes, but is not limited to, mu and combinations thereof.

[0046] In some embodiments, the platform semi-quantitatively measures gene expression levels. It is configured to determine, that is, rather than being measured in separate or absolute expressions. The expression levels are estimated values ​​and / or relative to each other, or identified markers ( For example, it is measured in comparison to the expression of another "standard" or "reference" gene.

[0047] In some embodiments, semi-quantitative measurements involve the estimation of genes within the signature or To provide relative expression levels, a signal indicating the identified mRNA is detected until... The PCR cycle is performed, and the number of PCR cycles required until detection is used is called "real This includes "time PCR".

[0048] Real-time PCR platforms include, for example, TaqMan® low-density It includes an array (TLDA), and in that TLDA, the sample is subjected to real-time PCR. On an array card having a collection of wells to be processed, multiplexed reverse transcription is performed, followed by real Undergo timed PCR. Kodani et al. 2011, J. Clin. See Microbiol. 49(6):2175-2182. Real-time PCR The form also includes, for example, the sample preparation of Biocartis Idylla (trademark). The technology from to the result includes, in that technology, cells are lysed, D NA / RNA is extracted, real-time PCR is performed, and the results are detected.

[0049] Magnetic resonance platforms include, for example, T2 Biosystems® T It includes 2-magnetic resonance (T2MR®) technology, and in that technology Molecular targets can be identified in biological samples without the need for purification.

[0050] The terms "array," "microarray," and "microarray" are interchangeable. Refers to the arrangement of a collection of nucleotide sequences presented on a support. Any type of array. However, this can be used in the methods provided herein. For example, the array is a slub Solid supports such as glass (solid-phase arrays), or semi-fluid materials such as nitrocellulose membranes. It may be on a support. The array may also be presented on beads, i.e., in a bead array. It's possible. These beads are typically microscopic and made of, for example, polystyrene. It is possible. The array may also be presented on nanoparticles, which may be, for example, gold. However, it can also be made from silver, palladium, or platinum. For example, gold nanoparticles Nanosphere Verigene (registered trademark) system using advanced technology See also. Magnetic nanoparticles can also be used. Other examples include nuclear magnetic resonance microcoils. It is born. The nucleotide sequence is DNA, RNA, or any permutation thereof (permu (For example, nucleotide analogs such as locked nucleic acid (LNA)) It is possible. In some embodiments, the nucleotide sequence is more than genomic DNA. Exons / introduction to detect gene expression of priced or mature RNA species It extends to the boundary. Nucleotide sequences also include gene-derived partial sequences, primers, and whole genes. Arrays, non-coding arrays, coding arrays, published arrays, known arrays, or novel arrays It is possible. The array can also include antibodies that specifically bind proteins or metabolites. It may include peptides, proteins, tissues, cells, chemicals, carbohydrates, and other compounds.

[0051] The array platform includes, for example, the TaqMan® low-power version mentioned above. Density array (TLDA) and Affymetrix® microarray plastic A foam insert is included.

[0052] Hybridization and multi-signal coding detector platforms include: For example, NanoString nCounter(registered trademark) technology ((for example (The color coding attached to the target-specific probe corresponding to the gene expression transcript of interest) Hybridization of barcodes is detected), and Luminex (registration (Trademark) xMAP (Registered Trademark) Technology (Microsphere beads are color-coded) and coated with a target-specific probe for detection (e.g., gene expression transcript). (and Illumina® BeadArray (microbeads) The optical fiber bundles are collected on a planar silica slide and used as a target for detection (e.g., It includes gene expression transcripts (coated with specific probes).

[0053] Nucleic acid mass spectrometry platforms include, for example, Ibis Biosciences P The lex-ID (registered trademark) detector is included, and in that detector, DNA mass spectrometry is performed. It is used to detect amplified DNA using mass profiles.

[0054] Thermal cycler platforms include, for example, FilmArray®. Multiplex PCR system (extracts and purifies nucleic acids from untreated samples and performs nested multiplex PCR) (to implement), and RainDrop Digital PCR System (microfluidic chip) This includes a PCR platform based on the droplets used.

[0055] The term "computer-readable medium" refers to a medium that stores information (e.g., data and instructions) and is computer-readable. This refers to any device or system used to supply power to a computer processor. Examples of readable media include DVDs, CDs, hard disk drives, magnetic tapes, and nets. Streaming media and applications on the network (e.g., smartphones) This includes, but is not limited to, servers for (those found on phones and tablets). Not done. In various embodiments, aspects of the present invention, including data structures and methods, It can be stored on computer-readable media. Processing and data can also be stored on a variety of devices. It can run on desktop and laptop computers, and its device types include This includes, but is not limited to, tablets and smartphones.

[0056] As used herein, the term "biological sample" refers to the method provided herein. It includes any sample that can be taken from the subject and contains genetic material that can be used. For example, biological samples may include peripheral blood samples. The term "peripheral blood sample" refers to a sample from the circulatory system. This refers to blood circulating in the body, or a sample of blood taken from that system. Other samples are... This may include samples taken from the airway, such as sputum, nasopharyngeal swabs, and nasopharyngeal lavage solutions. This includes, but is not limited to, biological samples. It also includes samples taken from the lower respiratory tract. This may include, but is not limited to, bronchoalveolar lavage fluid and tracheal aspirate. No. Biological samples may also include any combination of them.

[0057] The term "genetic material" refers to the storage of genetic information in the nucleus or mitochondria of cells in living organisms. This refers to the substance used for genetic material. Examples of genetic material include double-stranded and single-stranded DNA, cD This includes, but is 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. This refers to ligomer.

[0059] As used herein, the terms “to treat,” “treatment,” and “to treat” are defined as follows: , a disease or disorder or one or more symptoms resulting from the administration of one or more treatments It refers to a decrease or improvement in the severity, duration, and / or progression of the condition. Such terminology is , reduced replication of viruses or bacteria, or other effects on the target of viruses or bacteria This refers to a reduction in the transmission of an allergen to an organ or tissue, or to another target. Treatment also involves allergy This may include treatment for ARIs resulting from non-communicable diseases such as asthma.

[0060] The term "effective dose" refers to the amount of therapeutic agent sufficient to produce a physiological effect in a target. The term "responsiveness" refers to the level of gene expression of a gene in a subject, and the state in which the subject is affected by a virus. A virus, in response to being infected with a virus or bacteria, or suffering from a non-communicable disease. Subjects that are not infected with bacteria or do not have a non-communicable disease, or control subjects This refers to the change in comparison with the gene expression level of the aforementioned gene.

[0061] The term "appropriate treatment regimen" refers to the treatment regimen required to treat a particular disease or disorder. This refers to the standard of A. Such regimens have the ability to produce a curative effect in the disease state. In many cases, it is necessary to administer therapeutic agents to the target patient. For example, a patient with bacteremia The treatment agents used to treat it are antibiotics, which include penicillin, cephalosporins, It contains fluoroquinolones, tetracyclines, macrolides, and aminoglycosides. However, it is not limited to these. Treatment for subjects with viral respiratory infections. Therapies include oseltamivir, RNAi antivirals, inhaled ribavirin, and monoclonal antivirals. It contains body respigum, zanamivir, and neuraminidase blocking agents, but The present invention relates to treatment with antiviral agents or antibiotics that are not yet available. The present invention intends to be used to determine the appropriate treatment regimen, which also includes non-infectious treatment. This includes treatment for ARIs caused by disease, and such treatment may, for example, not limited to, anti-inflammatory drugs. Staminas, decongestants, anticholinergic nasal sprays, leukotriene inhibitors, mast cell inhibitors Allergy treatment including administration of drugs, steroid nasal sprays, etc.; and, not limited to, inhalation Luticosteroids, leukotriene modifiers, long-acting β-agonists, mixed inhalers (for example) fluticasone-salmeterol; budesonide-formoterol; mometasone-formoterol (e.g., lorazepam), theophylline, short-acting β-agonists, ipratropium, oral and IVF This includes asthma treatments such as intravascular corticosteroids and omalizumab.

[0062] Such regimens involve therapeutic agents that have the ability to reduce symptoms associated with the disease state. In many cases, it is necessary to administer the drug to the elephant. Examples of such therapeutic agents include NSAIDs. Acetaminophen, antihistamines, β-agonists, cough suppressants, or symptoms associated with the disease process This includes, but is not limited to, other drugs that reduce the condition.

[0063] Method for creating (training) classifiers This disclosure describes the creation of a classifier used in a method for determining the etiology of acute respiratory diseases in a subject. This provides a method (also called training 10) for determining the etiology of ARI in the subject with high accuracy. A gene expression-based classifier that can be used to identify and characterize is being developed. It is being emitted.

[0064] Therefore, as shown in Figure 1, one aspect of this disclosure relates to acute respiratory infections (AR). I) A method for producing classifiers, wherein (i) bacterial, viral, or non-infectious acute respiratory Obtaining biological samples (e.g., peripheral blood samples) from multiple subjects suffering from respiratory infections (ii) Optionally, prepare RNA (for example, a transcriptome) from the above sample. (iii) A step to isolate the total RNA (for preparation) (105, not shown in Figure 1); (iii) A number of genes (i.e., some or all of the genes expressed in the RNA) Step 110: (iv) (v) Based on the above results, bacterial ARI classifiers, viral ARI The process includes step 130 for creating classifiers or non-infectious disease classifiers, and consists of the following: Or it provides a way to essentially become one with them.

[0065] In some embodiments, the sample is not purified after collection. Furthermore, the sample may be purified to remove exogenous material before or after cell lysis. In some embodiments, the sample undergoes cell lysis and extraction of cell material, followed by nucleic acid isolation. , and / or reduction of large amounts of transcripts such as globin or ribosomal RNA It is then refined.

[0066] In some embodiments, measuring gene expression levels is performed as described in the transcription. Steps to prepare one or more microarrays using a primer; multiple primers - A step of measuring the transcriptome using -; analyzing and correcting batch differences. This may include steps.

[0067] In some embodiments, the method involves targeting the final gene for the generated classifier. Stroke, related weight (W n ), and upload thresholds to one or more databases This further includes step 140.

[0068] An example of how to create the aforementioned classifier is detailed in Figure 2. As shown in Figure 2, bacterial Biological data from a cohort of patients including ARI, viral ARI, or non-communicable diseases. The sample is for each condition (i.e., bacterial acute respiratory infection, viral acute respiratory infection, To develop gene expression-based classifiers for diseases (or non-infectious diseases) It is possible. Specifically, bacterial ARI classifiers are viral ARIs or non-infectious diseases. This is obtained to reliably identify individuals with bacterial ARI for any of the following: The Rustic ARI classifier distinguishes between bacterial ARIs or non-infectious diseases (NIs) and viral A. It is acquired to reliably identify those with RI. Non-infectious disease classifiers are bacterial It is created to improve the specificity of viral ARI classifiers. Next, bacterial ARI Signatures for classifiers, viral ARI classifiers, and non-infectious disease classifiers are ( For example, it can be constructed by applying a sparse logistic regression model.

[0069] Subsequently, these three classifiers, if necessary, determine the classifier based on the maximum membership probability. A single ARI classifier named after following a one-to-many scheme that specifies Bell It can be integrated into a classifier. See also Figure 5. The integrated ARI classifier can be used in several embodiments. And this can be verified using one-out cross-validation in the same group from which it was derived, and / Alternatively, in some embodiments, a sample derived from a subject suffering from a disease of a known etiology This can be validated using publicly available human gene expression datasets. For example, validation can be performed using: Publicly available human gene expression datasets (e.g., GSE6269, GSE420) Using 26, GSE40396, GSE20346, and / or GSE42834) The dataset may be implemented, and they may be from at least two clinical groups (bacterial ARI, The selection criteria include the inclusion of viral ARIs (or non-infectious diseases).

[0070] The classifiers are diseases of known etiology, namely bacterial ARI, viral ARI, or non- This can be validated in a standard set of samples derived from subjects suffering from infectious diseases.

[0071] The training methodology described herein can be used by those skilled in the art to detect different gene expression ( For example, it can be easily rewritten into an mRNA detection and quantification platform.

[0072] The methods and assays described herein are based on gene expression, for example, the direct measurement of RNA. Measurement of the derived material (e.g., cDNA), and RNA product (e.g., encoding This may be done through the measurement of proteins or peptides. Gene expression can be extracted and screened. Any method of processing may be used and is within the scope of this disclosure.

[0073] In some embodiments, the measurement involves the detection and quantification of mRNA in a sample (e.g.) For example, this includes semi-quantification. In some embodiments, the gene expression level is one or This is adjusted for multiple standard gene levels ("normalization"). As is explained, normalization gives the quantity or relative amount (for example, of expressed genes). This is done to eliminate the technical variability inherent in the rat form.

[0074] In some embodiments, mRNA detection and quantification are performed first by reverse transcription and / Alternatively, it may include an amplification step, such as RT-PCR, including quantitative RT-PCR. In several embodiments, detection and quantification are performed to determine whether a biological sample is present or not. It can be obtained based on purified, unamplified mRNA molecules. Direct detection and measurement of RNA molecules is possible. In terms of type, hybridization with complementary primers and / or labeled probes This includes traditional Northern blotting and surface-enhanced Raman spectroscopy. The method (SERS) (which involves the surface of a plasmon-activated metal structure with a gene-specific probe) A sample exposed to light is irradiated with a laser, and the change in optical frequency as the light is scattered is measured. This includes (including doing).

[0075] Similarly, the detection of RNA derivatives such as cDNA typically involves complementary primers and / or hybridization with a labeled probe. This includes high-density oligonucleotides. Rheotide probe arrays (e.g., solid macroarrays and bead arrays) or related Probe-hybridization methods, as well as relative and absolute values ​​of specific RNA molecules Includes real-time, digital, and endpoint PCR methods for comparative quantification. This may include amplification and detection based on a limelase chain reaction (PCR).

[0076] Additionally, sequencing-based methods can detect RNA or RNA-derived material at the RNA or RNA-derived material level. It can be used for detection and quantification. When applied to RNA, sequencing The sequencing method is called RNA-seq, and it obtains qualitative information about RNA molecules derived from the sample (in the sample). RNA, or the sequence or presence / absence of its related cDNA) and quantitative information (copy It provides both (number) and (number). For example, Wang et al. 2009 Nat. Rev. See Genet. 10(1):57-63. Another sequence-based method is genetics. Sequential Analysis of RNA Expression (SAGE) is a surrogate method for measuring the expression levels of RNA molecules. Use a cDNA "tag".

[0077] Furthermore, the use of a proprietary platform for mRNA detection and quantification is also a key aspect of this. These may be used to implement the disclosure method. These examples are CELLULAR RESE Molecular Indexing (trademark) developed by ARCH, INC. The incorporated Pixel® system and NanoString® system are included. echnologies nCounter gene expression system; Illumina Corporation mRNA-Seq, Tag-Profiling, BeadArray (trademark) technology G, and VeraCode, PrimeraDx's ICEPlex system, This is Affymetrix's QuantiGene 2.0 multiplex assay.

[0078] For example, RNA from whole blood of the subject is processed using PAXgene(trademark) RNA tubes. RNA preservation reagents such as PreAnalytiX, Valencia, and Calif. are used. RNA can be collected using the standard PAXgene™ or Versag ene(TM) (Gentra Systems, Inc., Minneapolis, M. RNA can be extracted using the inn. RNA extraction protocol. Versagene The (trademark) kit yields higher yields from PAXgene (trademark) RNA tubes. It produces high-quality RNA. After RNA extraction, GLOBINCIea is used to reduce whole blood globin levels. You may use the trademarks r(Ambion, Austin, Tex.). The method uses a bead-oligonucleotide construct to bind globin mRNA, In the inventors' experiments, they found that they could remove more than 90% of globin mRNA. (This technology allows for the removal of large quantities of irrelevant transcripts.) This can increase the sensitivity of assays, such as with the Chloarray platform.

[0079] RNA quality can be evaluated by several means. For example, RNA quality is Immediately after extraction, the results are evaluated using the Agilent 2100 Bioanalyzer. This analysis can be used to quantify RNA quality by determining RNA completeness (RNA). It provides the Integrity Number (RIN). Also, after globin reduction, The amount can be compared to a globin reduction standard. In addition, scaling factors and batches can be used. The ground can be evaluated after hybridization with a microarray. .

[0080] Real-time PCR can be used to rapidly identify gene expression from whole blood samples. For example, isolated RNA is reverse transcribed, then amplified, and the resulting dsDNA is... A nonspecific fluorescent dye inserted into it, or a hybrid of the probe with its complementary DNA target. Sequence-specific DNA labeled with a fluorescent reporter, enabling detection only after dilation. It can be detected in real time using lobes.

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

[0082] Expression levels are typically determined using methods routinely practiced by those skilled in the art, for a specific purpose. Depending on the rat form, the data is normalized after detection and quantification.

[0083] For mRNA detection and quantification, and matched normalization appropriate for the platform Regarding methodology, it is simply whether to use carefully selected and judged patient samples for the training method. For example, the cohorts described below can be used with the genes in three signatures in the classifier for the platform to derive appropriate weighting values (coefficients). These subject-samples can also be used to derive coefficients and cutoffs for tests performed using different mRNA detection and quantification platforms. In some embodiments, the individual categories of the classifier (i.e., bacterial ARI, viral ARI, non-infectious diseases) are formed from cohorts that include various such causes of it. For example, the bacterial ARI classifier is obtained from a cohort with bacterial infections from multiple genera and / or species of bacteria, the viral ARI classifier is obtained from a cohort with viral infections from multiple genera and / or species of viruses, and the non-infectious disease classifier is obtained from a cohort with non-infectious diseases due to multiple non-infectious causes. For example, see Table 8. In this way, each classifier obtained does not sense the underlying bacterial, viral,

[0084] and non-infectious causes. In some embodiments, some or all of the subjects with non-infectious etiologies in the cohort have symptoms consistent with respiratory infections. In some embodiments, the signature is identified by a model based on their ability to separate phenotypes during the training process using a selected set of patient samples

[0085] ​​​​​​​​​​、Known as sparse linear classification, it can be obtained using a supervised statistical approach. The result of training is the gene signature and classification coefficients for those three comparisons. Together, the signature and coefficients provide a classifier or predictor. Training can also be used to establish a threshold or cut-off value. The threshold or cut-off value can be adjusted to vary the test performance, e.g., test sensitivity and specificity. For example, the threshold for bacterial ARI can be intentionally lowered if necessary to increase the sensitivity of the test for bacterial infection.

[0086] In some embodiments, classifier generation includes (i) assigning weights to each of the normalized gene expression values, inputting the weights and expression values for each gene into a classifier (e.g., a linear regression classifier) equation, and determining a score for the result for each of a plurality of subjects, then (ii) determining the classification accuracy for each result across the plurality of subjects, and then (iii) repeatedly adjusting the weights until the classification accuracy is optimized. Genes having non-zero weights are included in each classifier.

[0087] In some embodiments, the classifier is a linear regression classifier and the generation includes converting the score of the classifier to a probability using a link function. As is known in the art, the link function specifies the link between the target / output of the model (e.g., the probability of bacterial infection) and the systematic component of the linear model (in this case, the combination of explanatory variables including the predictor). It shows how the expected value of the response is related to the linear predictor of the explanatory variables. ​

[0088] Classification method This disclosure concerns whether the respiratory illness a patient has is caused by a bacterial infection or a viral infection, To further provide methods for determining whether the cause is non-infectious or not, The method relies on the use of the classifier obtained as taught herein. a) A defined set of genes (i.e., one or more of its three signatures) a) the step of measuring the expression level of ( ); b) the technology used to perform the measurement Steps to normalize the gene expression levels for G; c) Take those values ​​and These are predefined weighted values ​​(coefficients) for each gene in each signature. Bacterial classifiers, virus classifiers, and / or non-infectious disease classifiers having (i.e., Steps to input into the predictor; d) Predetermine thresholds, cutoff values, and likelihood of infection. Steps to compare the confidence interval or range of values ​​with the classifier output; and optionally, e) classifier This may include a step of reporting the results together.

[0089] A simple overview of such a method is provided in Figure 3. In this representation, three genes Each of the sub-signatures represents a different ARI etiology (bacterial or viral) or infection. These signs do not provide information about the patient's host response to the disease state (NI). Necha is a bacterial ARI, viral ARI, or a diseased state without infection. In response to one of the three clinical conditions, a consistent and coordinated increase in expression levels or These are a group of gene transcripts that cause a decrease. These signatures were carefully determined. This is derived using a sample of patients with the condition of interest (training 10).

[0090] With respect to FIG. 3, after obtaining a biological sample (e.g., a blood sample) from a patient, in some embodiments, mRNA is extracted. The mRNA (or a defined region of each mRNA) is quantified for all or a subset of the genes in the signature. Depending on the instrument used for quantification, the mRNA may first have to be purified from the sample.

[0091] The signature is a reflection of the clinical state and is defined relative to at least one of the other two possibilities. For example, the bacterial ARI signature is a set of biomarkers (represented herein by gene mRNA transcripts) that distinguish patients with bacterial ARI from patients without bacterial ARI (including patients with viral ARI or, if applicable in this context, patients with non-infectious diseases). The viral ARI signature is defined by a set of biomarkers that distinguish patients with viral ARI from patients without viral ARI (including patients with either bacterial ARI or non-infectious diseases). The non-infectious disease signature is defined by a set of biomarkers that distinguish patients with non-infectious etiologies from patients with either bacterial or viral ARI. The normalized expression level of each gene in the signature (e.g., the first column of Table 9) is an explanatory or independent variable or feature used in the classifier. As an example, the classifier may have the general form as a probit regression formula:

[0092] P(having condition) = Φ(β1X1 + β2X2 + … + β X d X d (Equation 1) ​​​​​​​​​In the formula, the condition is bacterial ARI, viral ARI, or non-infectious disease; Φ(.) is The probit (or logistic, etc.) link function; {β1, β2, ..., β d} These are the coefficients obtained during training (for example, columns 2, 3, and 4 from Table 9). (The coefficients are also referred to as "weights" in this specification as {w1, w2, ..., w d} can be expressed as);{ X1, X2, ..., X d} is the normalized gene expression level of the signature; as well as d This is the size of the signature (i.e., the number of genes).

[0093] As is understood by those skilled in the art, the values ​​of the coefficients for each explanatory variable are given by probit times Used to measure the expression of a gene or subset of a gene used in a regression model. It varies depending on the technology platform. For example, Affymetrix U13 Regarding gene expression measured by 3A 2.0 microarrays, a classifier algorithm The coefficients for each of the features in are shown in Table 9.

[0094] The sensitivity, specificity, and overall accuracy of each classifier are determined using receiver operating characteristic (ROC) curves. Therefore, it can be optimized by changing the threshold for classification.

[0095] Another aspect of this disclosure is whether acute respiratory infections (ARIs) in the subject are of bacterial or viral origin. A method for determining whether an infection is of a biological or non-infectious origin, wherein a) the subject is biological Steps to obtain a sample, (b) the gene expression profile of the target gene from the biological sample, defined This is determined by evaluating the expression levels of the gene set (i.e., the three signatures). Steps to determine, c) the technology used to perform the above measurement regarding gene expression Step d) Normalize the level to obtain a normalized value, and then apply the normalized value to each signature. A bacterial classifier having a predefined weighting value (coefficient) for each gene in the given location. Steps to input into the virus classifier and the non-infectious disease classifier (i.e., predictor), e) The output of the classifier is defined as a range of thresholds, cutoff values, or values ​​indicating the likelihood of infection. Steps to compare with the surrounding area, and e) the sample is determined to be bacterial, viral, or non- This includes the step of classifying something as an infectious disease, and consists of, or is essentially derived from, that step. The method provides a way to achieve this. In some embodiments, the method shows the probability of the pathogenesis of ARI. This further includes the step of creating a report that assigns a score to the patient.

[0096] During training, the classifier developed using the training sample set diagnoses new individuals ("classification"). It is applied for the purpose of predicting the following: For each subject or patient, a biological sample is collected. Each of the genes identified by the signature found during the sample was selected and trained. The normalized expression level (i.e., relative mRNA expression level) is used as input for the classifier. Yes. The classifier also uses weighting coefficients discovered during training for each gene. Output The classifier is used to calculate three probability values. Each probability value is for bacterial AR I. The prognosis for three possible clinical conditions: viral ARI, and non-infectious disease. It can be used to make a decision.

[0097] In some embodiments, the result of each classifier, i.e., a new subject or The probability that a patient has bacterial ARI, viral ARI, or a non-infectious disease is reported. In its final form, three signatures with corresponding coefficients are applied to individual patients. Then, there are three probability values: bacterial ARI, viral ARI, and non-infectious disease. The probability of having such a value is obtained. In some embodiments, these values ​​are the confidence with which the classification is made. It may be reported in comparison to a reference range indicating the degree. In some embodiments, the classifier output is The classifier score or probability is compared to a threshold, for example, bacterial ARI, viral A If the threshold is exceeded to indicate the presence of RI or one or more non-communicable diseases, it is considered "positive". It can be reported. If the classifier score or probability cannot reach the threshold, the result The results are reported as "negative" for each of the conditions. Optionally, bacterial and viral Only values ​​for septic ARI were reported, and the report indicated that the person had the disease but was not infected. I will not comment on the possibility of it not happening.

[0098] A classifier obtained on one platform may exhibit optimal performance on another platform. It should be noted that this may not be the case. This is due to the promiscuous crossbreeding of probes, or plastic It could be due to other technical issues specific to the platform. Adapt a signature from a form, as taught herein, to another platform. Methods for achieving this are also described herein.

[0099] For example, the signature obtained from the Affymetrix platform is the signature About the gene and / or Affymetrix platform in NETHA For alternative genes correlated with the gene in the aforementioned signature, the corresponding TLDA The use of lobes can be adapted to the TLDA platform. Table 1 shows Affyme trix probes and the genes they measure, plus TLDA for technical reasons. A may not function well on the platform, or there may be no related TLDA probes. Either to replace the ffymetrix probe or for other reasons regarding the gene probe This is a list of "replacement genes" that are introduced as replacements. These replacements are high It may show correlated genes, or bind to different positions in the same gene transcript. These may be probes. Additional genes, such as panviral gene probes, may be included. (Table 1) The weights shown are calculated for classifiers run on the microarray platform. These are the estimated weights. Weights that have not been estimated are indicated by "NA" in the table (see below). Example 4 provides a complete rewrite of these classifiers to the TLDA platform. (to do so). A reference probe for TLDAs (i.e., a normalized gene, e.g., TRAP) 1. PPIB, GAPDH, and 18S) also use weights and Affymetrix In the column for the lobeset ID (these are not the classifier part), it is indicated as "NA". This may result in additional legacy features that do not necessarily correspond to the Affymetrix probe set. The probe probe also shows "NA" in the Affymetrix probe set ID column. It will be done.

[0100] JPEG0007849855000001.jpg233146JPEG0007849855000002.jpg240129JPEG0007849855000003.jpg239124JPEG0007849855000004.jpg93133

[0101] Further consideration of this signature example for the TLDA platform can be found in the implementation below. Examples 3 and 4 are provided.

[0102] This method for determining the etiology of ARIs can be combined with other tests. For example, if the patient If it is determined that the patient has viral ARI, follow-up testing will be performed. Whether it is possible to directly detect A or B, or whether accommodations show such infections. This could involve determining whether the main response can be detected. Similarly, bacterial AR Follow-up tests on the results of I directly detect Gram-positive or Gram-negative bacteria. Whether it is possible, or whether a host response indicating such an infection can be detected. It may be determined. In some embodiments, a classifier is used to classify the class of infection. To determine, and also, using pathogen-specific probes or detection methods, a specific pathogen Simultaneous tests can be conducted to examine the body. For example, Eley et al.'s U.S. Patent Publication Publication No. 2015 / 0284780 (Method for detecting active tuberculosis); Tsalik et al. See U.S. Patent Publication No. 2014 / 0323391 (Method for Classifying Bacterial Infections).

[0103] Method for determining secondary classification of ARI in a subject This disclosure also provides a method for classifying objects using a secondary classification scheme. Another aspect of the present invention relates to acute respiratory infections (ARIs) in subjects that are of bacterial origin or are viral. A method for determining whether the origin is Rus or non-infectious, wherein (a) the organism (b) Obtain a biological sample, and determine the gene expression profile of the target gene from the biological sample. By evaluating the expression levels of the pre-defined gene set (i.e., the three signatures), (c) The steps to determine the technology used to perform the measurement (d) Each sig A classification system in which each gene in a gene has a defined weight (coefficient). (e) the step of inputting the normalized value to the child (i.e., the predictor) Steps that compare against a predefined threshold, cutoff value, or range of values ​​indicating the likelihood of infection. (f) If the sample is negative for bacteria, the virus classifier and non-infectious The step of repeating step (d) using only the classifier, and (g) the sample This includes the step of classifying whether it is an infectious disease or a non-infectious disease, and consists of these steps, Or it provides a way to essentially become one with them.

[0104] Another aspect of this disclosure is that acute respiratory infections (ARIs) in the subject are of bacterial origin or are viral. A method for determining whether the origin is Rus or non-infectious, wherein (a) the organism (b) Obtain a biological sample, and determine the gene expression profile of the target gene from the biological sample. By evaluating the expression levels of the pre-defined gene set (i.e., the three signatures), (c) the steps to determine the technology used to perform the measurement, (d) Normalizing the gene expression level to obtain a normalized value in each signature A classifier having a predefined weighting value (coefficient) for each gene (i.e., (e) input the normalized value to the predictor, and (f) set the output of the classifier to a defined threshold (f) the step of comparing the value, cutoff value, or range of values ​​indicating the likelihood of infection If the sample is negative for the virus, only bacterial and non-infectious classifiers are used. Then, the step (d) is repeated, and (g) the sample is found to be a bacterial pathogen or This includes the step of classifying them as non-infectious diseases, and consists of, or is essentially derived from them. Provides a way to become a target.

[0105] Another aspect of this disclosure relates to whether acute respiratory infections (ARIs) in the subject are of bacterial origin or , a method for determining whether the origin is viral or non-infectious, wherein (a) the subject (b) Obtain a biological sample, and from the biological sample, obtain the gene expression profile of the target gene. This evaluates the expression levels of a predefined set of genes (i.e., three signatures). (c) The steps determined by the above, and the technology used to perform the above measurement. (d) normalize the gene expression level to obtain a normalized value for each signature. A classifier that has a predefined weighting value (coefficient) for each of the genes in (e) The steps of inputting the normalized value to the predictor, and (f) the output of the classifier, Steps include comparing the value to a threshold, cutoff value, or range of values ​​indicating the likelihood of infection, (f ) If the sample is negative for non-infectious diseases, the virus classifier and bacterial classification Using only the child, repeat step (d), and (g) the sample with a will The process includes the step of classifying whether the cause is uric acid or bacterial, and consists of or It provides a way to essentially become one with them.

[0106] In some embodiments, the method assigns patients a score indicating the probability of ARI etiology. This further includes the step of creating a report that will hit the target.

[0107] The classification of a patient's condition using a secondary classification scheme is shown in Figure 4. This example In this context, the bacterial ARI classifier classifies patients with bacterial ARI as those without bacterial ARI. This distinguishes it from patients with other causes (which may instead be viral ARI or non-infectious). Subsequently, a secondary classification was imposed on patients with nonbacterial ARI, distinguishing between viral ARI and non-infectious ARI. Further distinction can be made between diseases. The same process of primary and secondary classification is also possible. It can be applied to the Rustic ARI classifier, in which case it is determined that there is no viral infection. Patients who have been diagnosed are subsequently secondarily classified as having bacterial ARI or a non-infectious etiology. Similarly, by applying a non-infectious disease classifier as the primary test, the patient may have such a non It is determined whether the patient has an infectious disease or, instead, symptoms of an infectious cause. Secondary The classification step is whether the infectivity is caused by a bacterial pathogen or a viral pathogen. To decide.

[0108] The results from the three primary and three secondary classifications mentioned above can be interpreted by those skilled in the art using various techniques. These are then totaled (for example, sum, number, or average) to create a practical report about the provider. It can be done. In some embodiments, this secondary level of classification is used The genes may be some or all of the genes listed in Table 2.

[0109] In such examples, the three classifiers mentioned above (bacterial classifier, viral classifier, and non- The infectious disease classifier is used to perform the first level of classification. Subsequently, nonbacterial For patients with sexually transmitted infections, the secondary classifier is viral ARI, while the non-infectious disease is also classified as such. It is defined to distinguish it from other things (Figure 4, left panel). Similarly, nonviral infections For patients with this condition, a new classifier is used to distinguish viral illnesses from non-infectious diseases. (Figure 4, central panel), and in the first step, they were classified as having a non-communicable disease. For patients who do not have a diagnosis, a new classifier has been developed to distinguish between viral and bacterial ARIs. It is used (Figure 4, right panel).

[0110] In this two-stage method, nine probabilities can arise, and these probabilities can be combined in several ways. This is possible. Each has a probability of being a bacterial ARI, a viral ARI, and a non-infectious disease. Two strategies are described herein as methods for harmonizing the three sets of predictions that we possess. For example, the best predicted mean probability: the average of all predicted probabilities for bacterial ARI. The average value was calculated, along with all predicted probabilities for viral ARIs, and all predicted probabilities for non-infectious diseases. The same applies to probabilities. The highest average probability represents the diagnosis.

[0111] Maximum number of predictions: Instead of averaging the predicted probabilities for each condition, for that patient sample, a specific diagnosis The number of times when a diagnosis (i.e., bacterial ARI, viral ARI, or non-infectious disease) is predicted. The number is coefficientd. The best scenario is that the three classification schemes show the same answer (for example, S For Keem 1, bacterial ARI; for Scheme 2, bacterial ARI; and for Scheme 3, This is the case with bacterial ARI. In the worst case, each scheme will specify a different diagnosis, and 3 This refers to creating a tie in scoring.

[0112] The results of stage 1 separation using the previously described patient sample training set are shown, for example, in Table 3 (columns). The clinical classification presented is similar to that presented in the row (diagnostic test predictions). I got it.

[0113] JPEG0007849855000005.jpg32167

[0114] After classifying into two sections using the best predictive mean probability strategy, the results are shown in Table 4 (columns). This may be similar to the bed classification (diagnostic test prediction presented in the row).

[0115] JPEG0007849855000006.jpg31168

[0116] After a two-stage classification using the strategy with the highest number of predictions, the results are shown in Table 5 (Clinical parts presented in the columns). Similar to the diagnostic test predictions presented in the row.

[0117] JPEG0007849855000007.jpg33168

[0118] For example, as described above and / or summarized in Table 2, classification can be achieved. This is possible. Table 2 shows the results for three separate classification strategies in which different diagnostic questions are answered. Summarize the gene membership. A total of 270 genes, including three complex classifiers. There is a probe. The first is the same as the one presented in Example 1 below, BVS (bacterial These are called ARIs (viral ARIs, SIRS). These probes are shown in Table 9. This is the same as the one described, and Table 9 provides the probe / gene weights used for classification. These also correspond to the genes presented in Table 10.

[0119] The second is called 2L, which is two layers or two stages. This is the hierarchical gap shown in Figure 4. It is a room.

[0120] The third is similar to BVS, but also includes a healthy control population (as also described in Example 1). This group includes a one-stage classification scheme, BVSH. This group has insufficient controls for non-infection. While it has been shown that there are use cases where differentiation from health may be clinically important, there are also use cases where this may be necessary. For example, this could include a continuous measurement of signatures that correlate with the recovery period. Used to identify patients who have been exposed to a chromosomal substance and who are experiencing prodromal symptoms rather than being asymptomatic. It is possible. In the BVSH scheme, four groups were selected in the training cohort, with bacterial ARI, The data shows groups with viral ARI, SIRS (non-infectious disease), and healthy individuals. These four groups have four distinct signatures that distinguish each class from all other possibilities. It is used to produce [something].

[0121] Table 2 Legend: Probe = Affymetrix probe ID BVS = Bacterial ARI, Viral ARI, and Non-Infectious Diseases (with Respiratory Symptoms) Three classifier models trained in patients with [condition]. 1 is that this probe [performs] these three [conditions]. This indicates that the probe is included in the classifier model. 0 means that the probe exists in this classification scheme. It indicates that something is not present. BVS-BO = A gene included in the bacterial ARI classifier as part of the BVS classification scheme. This classifier identifies patients with bacterial ARIs as having other etiologies (viral ARIs). Alternatively, identify specifically from 10). BVS-VO = This column identifies genes included in the viral ARI classifier. And, similar to BVS-BO. This classifier classifies patients with viral ARIs as other It is specifically identified based on its etiology (bacterial ARI or non-infectious disease). BVS-SO = Except for identifying genes included in the non-infectious disease classifier. , similar to BVS-BO or BVS-VO. This classifier has a non-infectious disease. To specifically identify the patient from other etiologies (bacterial or viral ARIs). 2L refers to a two-tiered hierarchical classification scheme. In this column, 1 is the identified probe. This indicates that the gene was included in the classification task. This two-stage classification scheme is itself It consists of three distinct, step-by-step tasks. The first step involves applying one-to-other, with one being a bacterial AR. It may be a viral ARI, or a non-infectious disease. The specified subject falls under the "Other" category. If it falls into the - category, a second stage of classification occurs to distinguish it from the remaining possibilities. 2L-SO is, The first stage of the model for determining whether a given subject has a non-communicable disease, Subsequently, SL-BV, which can distinguish between bacterial and viral ARIs, was developed as a possibility. In these columns, 1 indicates that the gene or probe is in its identified classification model. This indicates inclusion. 2L-BO and 2L-VS constitute another two-tiered classification scheme. 2L-VO and 2L-SB include a third model in the two-stage classification scheme. Finally, BVSH included healthy individuals in the training cohort, and therefore bacterial ARI, A classifier for viral ARIs or healthy status compared to non-infectious diseases. This refers to the Bell classification scheme. The dark gray BVSH column represents any of the classifications included in this scheme. Identify the gene or probe. This scheme itself is "1" in these columns. BVSH-BO and BVSH, which have the respective probe / gene compositions shown. It consists of -VO, BVSH-SO, and BVSH-HO.

[0122] Table 2 shows the virus classifiers, bacterial classifiers, and that are constructed according to the required tasks. This provides an overview of the use of members of the gene set for non-infectious disease classifiers. This indicates the membership of that gene in the classifier.

[0123] JPEG0007849855000008.jpg237135JPEG0007849855000009.jpg234146 JPEG0007849855000010.jpg236142JPEG0007849855000011.jpg235144JPEG0007849855000012.jpg235148 JPEG0007849855000013.jpg234146JPEG0007849855000014.jpg235148 JPEG0007849855000015.jpg236143JPEG0007849855000016.jpg234147JPEG0007849855000017.jpg236147JPEG0007849855000018.jpg235146 JPEG0007849855000019.jpg235127JPEG0007849855000020.jpg234147JPEG0007849855000021.jpg235142JPEG0007849855000022.jpg234145 JPEG0007849855000023.jpg234148JPEG0007849855000024.jpg23584

[0124] Method for treating subjects with ARI Another aspect of this disclosure relates to treating acute respiratory infections (ARIs) of unknown etiology in the subject. A method for placing (a) a biological sample from an object, (b) a biological sample From the target gene expression profile, define a set of genes (e.g., one, two, Alternatively, it is determined by evaluating the expression levels of three or more signatures. Step (c) Regarding the technology used to perform the measurement, as necessary (d) normalize the gene expression level to obtain a normalized value for each signature. A bacterial taxonomicon having a predefined weighting value (coefficient) for each of its genes, The normalized values ​​are input to the virus classifier and the non-infectious disease classifier (i.e., the predictor). (e) the output of the classifier is set to a predefined threshold, cutoff value, or infection threshold (f) A step of comparing the sample with a range of values ​​indicating the expected value, and a step of determining whether the sample is bacterial, viral, or The steps of (g) classifying it as a pathogenic or non-infectious disease, and in step (f) The step of applying an appropriate treatment regimen to the subject as identified therein, To provide a way of being composed of, or essentially composed of, them.

[0125] In some embodiments, step (g) determines that the etiology of ARI is bacterial. If so, this includes administering antibacterial treatment. In other embodiments, step (g) is If the etiology of ARI is determined to be viral, this includes administering antiviral treatment.

[0126] After the etiology of the ARI in question was determined, she decided on the treatment, for example, if the ARI was viral. If it is determined that she has been diagnosed, she may receive antiviral treatment, and / or she may have a course of infection. She may be isolated in her home during this time. Alternatively, if it is determined that the ARI is bacterial, A bacterial treatment regimen may be administered (e.g., antibiotic administration). It is classified as a non-infectious disease. Those who are diagnosed may be sent home, or may receive further diagnosis and treatment (e.g., allergies, asthma). They may be examined for reasons such as breathing problems.

[0127] However, the laboratory aims to identify the etiology of ARI and administer appropriate treatment. The purpose is to be able to communicate the gene expression levels of classifiers to physicians, so peripheral blood tests Those who have administered the treatment do not need to perform a comparison. In addition, after the doctor examines the patient, Peripheral blood samples are collected, the samples are assayed for a classifier, and the patient's etiological condition is sent to the agent. The plan is to instruct the agent to report the condition to the doctor. The doctor will investigate the etiology of ARI. Upon obtaining the diagnosis, a physician can order appropriate treatment and / or isolation.

[0128] The methods provided herein are intended to properly treat patients and reduce the inappropriate use of antibiotics. Therefore, it can be effectively used to diagnose the pathogenesis of diseases. Furthermore, as presented herein The method provided has various other uses, including (1) exposure to pathogens, Furthermore, it is possible to detect individuals with a disease that is not symptomatic, but is likely to occur at any moment, based on the host. The study (for example, in a scenario of natural disease transmission through a population, however, (2) In the case of o-terrorism, (3) in the setting of a clinical trial, or in the setting of an immune population For monitoring purposes, either for monitoring a response to a vaccine or drug. (3) For host-based testing, screening for imminent disease before placement. Host-based testing (for example, military deployment or civilian boarding such as a cruise ship) (in Nario), and (4) ARI in livestock (e.g., avian influenza and other p Host-based screening for viruses that may cause an epidemic. This includes, but is not limited to, exams.

[0129] Another aspect of this disclosure includes (a) means for extracting biological samples; and (b) teaching the foregoing. Multiple synthetic oligonucleotides having regions homologous to a group of gene transcripts as shown. Means for preparing one or more arrays consisting of ocides; and (c) instructions for use. Acute respiratory sensation in an object, including books, consisting of them, or essentially consisting of them. We provide a kit for determining the etiology of infectious diseases (ARIs).

[0130] Another aspect of this disclosure relates to a kit for evaluating acute respiratory infection (ARI) classifiers. A method using (a) a group of gene transcription products as taught herein One or more arrays consisting of multiple synthetic oligonucleotides having regions homologous to the substance. (b) A step of preparing the normalized gene; (b) an oligonucleotide having a region homologous to the normalized gene. Steps to add to the array; (c) from subjects suffering from acute respiratory infection (ARI) (d) Steps to obtain a physical sample; (d) Isolate RNA from the sample and transcribe it (e) Steps to prepare the transcriptome; (e) Steps to measure the transcriptome on the array. (f) Normalize the measured values ​​of the transcriptome against the normalized gene and classify them. Steps to electronically transfer normalized measurements to a computer in order to execute the child algorithm. (g) the step of preparing a report, and optionally, (h) appropriate action based on the results. A method that includes, consists of, or is essentially derived from, the steps of applying a solution. ru.

[0131] Classification system Regarding Figure 11, the classification system and / or computer program product 1100 This may be done in the platform according to the various embodiments described herein, or Therefore, it can be used in classification systems and / or computer program products. 1100 is any appropriate software, firmware, and / or hardware It can operate to receive, transfer, process, and store data using various combinations. It may exist and may be standalone, as well as / or known as the internet Any ordinary public and / or including all or part of the global communications network This is achieved through private, real and / or virtual, wired and / or wireless networks. One or more enterprises, applications, personals, widespread use, and can be interconnected. It can be embodied as a and / or embedded computer system, and in various forms of tangible expression. This may include non-transient computer-readable media.

[0132] As shown in Figure 11, the classification system 1100 is one or more operators A running system and / or one or more applications It may include a processor subsystem 1140 containing multiple central processing units (CPUs). Although one processor 1140 is shown, there may be multiple processors 1140, and they They can be either electrically interconnected or independent. It should be understood that the processor 1140 receives data from memory devices such as memory 1150. Executing computer program code to perform some of the operations and methods described herein It is configured to perform at least part of the function, and any ordinary or special-purpose processor The processor includes a digital signal processor (DSP) and a field programmer. Multi-gate arrays (FPGAs), special-purpose integrated circuits (ASICs), and multi-core This includes, but is not limited to, processors.

[0133] The memory subsystem 1150 is random access memory (RAM), read-only Memory (ROM), Erasable Programmable Read-Only Memory (EPROM), or This is a hierarchy of memory devices, such as flash memory and / or any other solid-state memory devices. It may include.

[0134] A memory circuit 1170 may also be provided, which may include, for example, a portable computer disk Hard disks, portable compact disks, read-only memory (CD-ROM) Optical storage devices, magnetic storage devices, and / or any other type of disk or tape A memory subsystem based on the .mechanism may be included. The memory circuit 1170 is a classification system 11 It can provide non-volatile storage for data / parameters / classifiers about 00. Memory circuit 11 70 may include disk drives and / or network storage components. The memory circuit 1170 stores code and / or other data that can be executed by the processor 1140. It can be used to store data that can be accessed. In some embodiments, The memory circuit 1170 is used in the classification system 1110, which includes signatures, weights, thresholds, etc. It can store a database that provides access to data / parameters / classifiers. Any combination of multiple computer-readable media can be utilized by the memory circuit 1170. Computer-readable media are computer-readable signal media or computer-readable storage media. It can be a body. Computer-readable storage media include, for example, electronic, magnetic, optical, electromagnetic, infrared A system, device, or apparatus of wires or semiconductors, or any suitable combination thereof. Possible, but not limited to those. More specific examples of computer-readable storage media (inclusive) (Not a target list) includes portable computer diskettes, hard drives, random Access memory (RAM), read-only memory (ROM), erasable programmable Read-only memory (EPROM or flash memory), portable compact disc Creed-only memory (CD-ROM), optical storage device, magnetic storage device, or the aforementioned Any suitable combination of the following is included. When used herein, computer-readable storage media The body is used by, or connected to, a command execution system, device, or apparatus. It may be any tangible medium capable of containing or storing the program.

[0135] The input / output circuit 1160 includes a display, and / or a keyboard, touchscreen. This may include a clean and / or user input device such as a pointing device. The device attached to the input / output circuit 1160 is used by the user of the classification system 1100. It can be used to supply information to the processor 1140. It is attached to the input / output circuit 1160. The devices included are networking or communication control devices and input devices (keyboard, mouse). This includes (such as touchscreens) and output devices (printers or displays). The input / output circuit 1160 also provides a display and / or printer. It can provide an interface to the system, and the result of the operation of the classification system 1100 is the classification system This can be communicated to those devices so as to be provided to users of the Mu1100.

[0136] The optional update circuit 1180 provides updates to the classification system 1100. It may be included as an interface for this purpose. The update will include memory 1150 and / or Or the code that is stored in the memory circuit 1170 and executed by the processor 1140 This may include updates. Updates provided via update circuit 1180 are Furthermore, data that maintains information about the classification system 1100, such as signatures, weights, and thresholds, is also used. Up portion of the storage circuit 1170 related to database and / or other data storage formats May include updates.

[0137] The sample input circuit 1110 of the classification system 1100 is a plug as described above. Toform can provide an interface for receiving biological samples to be analyzed. The sample input circuit 1110 receives biological samples supplied by the user to the classification system 1100. A classification system 1100 and / or platform to receive and process scientific samples. It may include mechanical and electrical elements for transporting biological samples within the system. Sample input circuit 1 110 identifies barcode-labeled containers for the identification of samples and / or test order forms. It may include a barcode reader. The sample processing circuit 1120 is for the automated analysis of biological samples. To prepare the material, the biological classification system 1100 and / or platform The sample can be further processed. The sample analysis circuit 1130 automatically processes the processed biological sample. The sample analysis circuit 1130 analyzes the biological samples supplied to the classification system 1100. Regarding the material, for example, it is used to measure the gene expression levels of a defined set of genes. The sample analysis circuit 1130 can also normalize by normalizing gene expression levels. Gene expression levels may be generated. The sample analysis circuit 1130 receives information from the memory circuit 1170 regarding bacterial acute Respiratory tract infection (ARI) classifiers, viral ARI classifiers, and non-infectious disease classifiers These classifiers can be extracted and used to define each gene in a defined set of genes. Includes predefined weighted values ​​(i.e., coefficients). Sample analysis circuit 1130 is used for bacterial acute respiratory Select from the tract infection (ARI) classifier, viral ARI classifier, and non-infectious disease classifier. Normalized gene expression values ​​can be input to one or more selected acute respiratory disease classifiers. The analysis circuit 1130 is used to analyze bacterial ARI, viral ARI, and one of the non-infectious diseases. The probability of etiology for multiple conditions is calculated based on the classifier, and acute respiratory disease in the subject is The determination of whether it is of bacterial origin, viral origin, non-infectious origin, or some combination of these. A constant output can be controlled via the input / output circuit 1160.

[0138] Sample input circuit 1110, sample processing circuit 1120, sample analysis circuit 1130, input / output circuit The path 1160, the memory circuit 1170, and / or the update circuit 1180 are at least It is also partially executed under the control of one or more processors 1140 of the classification system 1100. It is possible. As used herein, "running under the control" of processor 1140 means Sample input circuit 1110, sample processing circuit 1120, sample analysis circuit 1130, input / output circuit This is carried out by the memory circuit 1170 and / or the update circuit 1180. The operation is performed and / or instructed by processor 1140, at least in part. However, at least part of the operation of those components is independently electrical or mechanical. This means that automation is not ruled out. Processor 1140 is a computer By executing the program code, the classification system 11 as described herein The operation of 00 can be controlled.

[0139] Computer program code for operating in accordance with the aspects of this disclosure is an object Script-oriented programming languages, such as Java®, Scala, and SmallTal. k, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python Traditional procedural programming languages ​​such as "C" and V isual Basic, Fortran 2003, Perl, COBOL 2002 PHP, ABAP, dynamic programming languages, such as Python, Ruby One or more programming languages, including y and Groovy, or other programming languages. It can be written in any combination of programming languages. The program code is completely classified by the system. On the M1100, a portion is on the classification system 1100, standalone software package As a cage, partly on the classification system 1100 and partly on a remote computer It can be run on a remote computer or server, or entirely remotely. In this situation, the remote computer is on a local area network (LAN) or Classification system 110 through any type of network, including wide area networks (WANs) It may be connected to 0, or that connection may be (for example, by using an Internet Service Provider) (via the internet) to an external computer or cloud computing environment This may be done in or under Software as a Service (SaaS) It can also be offered as a service such as the following.

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

[0141] In some embodiments, the system allows the user to test the biological sample to be tested. Simply insert it, and then after a while (preferably a short time, for example, 30 minutes or 45 minutes, or 1 hour, 2 hours, or 3 hours, up to a maximum of 8 or 12 hours. You can receive the system's output results for up to 24 or 48 hours. It is a system with integrated components, from sample preparation to results.

[0142] In its application, the present invention is as shown in the following description or in the following drawings: It should be understood that this is not limited to the details of the build or the placement of components. The invention is capable of other embodiments and can be implemented or performed in various ways. .

[0143] The enumeration of value ranges in this specification does not apply unless otherwise indicated herein. It serves as a concise way to refer to each individual value that goes inside. This is intended to be the case, and each distinct value is as if it were individually listed in this specification. As such, it is incorporated into this specification. All methods described herein are as described herein. Unless otherwise specified in the details or clearly contradicted by the context, , can be carried out in any appropriate order. All examples provided herein, The use of illustrative language (e.g., "such as") further indicates that the present invention is more... This is merely intended to clarify, and unless otherwise asserted, the present invention This does not imply any limitation to the scope of [the term]. Any words used herein are not representative of any claim. Elements that are not included should not be interpreted as being essential for carrying out the present invention.

[0144] Any number range listed herein includes all values ​​from the lower limit to the upper limit. It is also understood that: For example, if the concentration range is stated as 1% to 50%, then 2 Values ​​such as %~40%, 10%~30%, or 1%~3% are explicitly stated in this specification. It is intended to be counted. These are merely examples of what is specifically intended, and most All possible combinations of numerical values ​​between the lowest and highest values ​​(including those values) are included in this application. It should be considered that this is clearly stated.

[0145] The following examples are illustrative only and are not intended to limit the scope. . [Examples]

[0146] [Example 1] Host gene expression classifiers diagnose the pathogenesis of acute respiratory diseases. Acute respiratory infections caused by bacterial or viral pathogens are the most common hospitalizations. This is one of the reasons why. Current pathogen-based diagnostic approaches are unreliable, and This was not timely, and therefore most patients received inappropriate antibiotics. Host response biomarkers are an alternative diagnostic approach to direct the use of antimicrobial agents. We will provide the chi.

[0147] The inventors have found that in emergency medical settings, host gene expression patterns are not the cause of disease infectivity. The question was whether to identify based on the cause of infection. Among those with acute respiratory infections, the present invention Those who determine whether an infectious disease is caused by a viral pathogen or a bacterial pathogen. It was decided.

[0148] The samples that formed the basis of the discovery were collected from the emergency departments of four tertiary care hospitals and a student health care facility. This was derived from an observational cohort study conducted with 44 healthy controls and community-acquired emergency patients. 273 patients with sexually transmitted respiratory infections or non-communicable diseases, suspected of having sepsis We selected from a larger cohort of patients (CAPSOD study). The mean age was 45 years. 45% of the participants were male. Further demographic information is incorporated herein by reference. Tsalik et al., (2016) Sci Transl Med 9(3 22): This can be found in Table 1 of 1-9.

[0149] Clinical phenotypes were determined by manual chart review. For respiratory virus pathogens, routine microbiological tests and multiplex PCR were performed. Peripheral whole blood gene expression was measured using a micro array. Sparsity logistic regression was used to develop classifiers for bacterial versus viral versus non-infectious diseases. Five independently derived datasets, including 328 individuals, were used for validation.

[0150] Gene expression-based classifiers were developed for bacterial acute respiratory infections (71 probes), viral acute respiratory infections (33 probes), or non-infectious etiologies (26 probes ). These three classifiers were applied to 273 patients, and class assignments were determined by the highest predicted probability. The overall accuracy was 87% (238 / 273 in agreement with clinical determination ), which was more accurate than procalcitonin (78%, p<0.03), and three published classifiers for bacterial versus viral infections (78 - 83%). The classifiers developed herein were externally validated in five publicly available datasets (AU C 0.90 - 0.99). The inventors compared the classification accuracy of tests based on host gene expression to procalcitonin and clinically determined diagnoses, including bacterial or viral acute respiratory infections or non-infectious diseases. The host peripheral blood gene expression response to infection provides a diagnostic strategy complementary to currently used diagnoses.

[0151] This strategy has been successful in characterizing the host response to viral ARI and bacterial ARI. These advances 8 . 8~13 and bacterial 11,14 ARI Nevertheless, their use as diagnostic tools in setting patient care presents several problems. This is hindering the development of host-based molecular signatures. The problem is that they are developed for the intended use group. 15 However... However, almost all ARI classifiers based on published gene expression are used as controls for healthy individuals. It uses individuals and focuses on small or uniform populations, and therefore, the affected It is not optimized for use in emergency treatment settings where patients exhibiting symptoms are not distinguished from each other. Furthermore, statistical methods used to identify gene expression classifiers include clustering, These strata often contain duplicated genes based on univariate testing or pathway association. Teji identifies related ecosystems, but does not fully utilize its diagnostic capabilities. Alternative methods, such as those exemplified, combine genes from unrelated pathways to achieve a more... The goal is to create a classifier that provides information.

[0152] method Classifier-induced cohort The research was approved by the relevant institutional review board and complied with the Declaration of Helsinki. All subjects The test subjects or their legally authorized representatives must provide written informed consent. It was served.

[0153] Community-acquired pneumonia and sepsis outcome diagnostic study (Clinical trial identification number NCT00258869) As part of ) Duke University Medical Center (D Emergency Department, UMC (Durham, NC), Durham VA Medical Center ter (DVAMC; Durham, NC), or Henry Ford Hospital In tal (Detroit, MI), patients suspected of having a community-acquired infection were Registered 16~19 As part of the community-acquired pneumonia and sepsis research, UNC Healthcare Emergency Department Additional patients were enrolled through (UNC; Chapel Hill, NC). This applies if you have a suspected infection and two or more systemic inflammatory response syndromes (SIRS). When the diagnostic criteria were presented, the patient was eligible. 20 ARI cases are handled by the Department of Emergency Medicine (SWG). , upper or lower respiratory tract (EBQ) or infectious disease (ELT) as determined by a physician This included patients with urinary symptoms. The determination was made at least 28 days after registration and previously published. Prior to categorization based on arbitrary gene expression using the defined diagnostic criteria, retrospectively performed Based on thorough manual review of medical records 17 The following were used to support these judgments. The complete information would likely not have been available to clinicians at the time of those evaluations. Pharyngitis Microbiologically confirmed bacterial ARIs included 4 people with the condition and 66 people with pneumonia. Seventy patients were identified. The microbiological etiology was normal culture of blood or respiratory samples. Nutrition, using urine antigen tests (Streptococcus or Legionella), or serological tests (Mycobacterium) It was determined by plasma. Patients with viral ARI (n=115) were viral Confirmation was made based on the identification of the etiology and the matching symptoms. In addition, using the same diagnostic method... Duke, who has confirmed viral ARI, as part of the DARPA Health Disease Prediction Study This included 48 students from the university. Regarding the identification of viral etiologies. The clinical trial involved a ResPlex II v2.0 viral PCR multiplex assay (Qiage Reinforced by Hilden, Germany. This panel was developed for influenza. Type A and Type B, adenovirus (B, E), parainfluenza types 1-4, respiratory multinucleus Human somatic virus types A and B, human metapneumovirus, human rhinovirus, coronavirus Viruses (229E, OC43, NL63, HKU1), Coxsackievirus / Echovirus It also detects the Boca virus. The results indicate that some of the registered patients have non-communicable diseases. It was decided that (n=88) (Table 8). An alternative diagnosis was established, and any routinely indicated Only if the results of the microbiological tests presented do not support an infectious etiology is it classified as a "non-infectious disease." The diagnosis of "illness" was made. Finally, healthy control (n) was selected from among the healthy volunteers who showed no symptoms. =44; median age 30 years; range 23-59 years) indicates the effect of aspirin on platelet function. Registered as part of a study on the subject, gene expression analysis was performed at a time before aspirin exposure. carried out 21 .

[0154] Procalcitonin measurement Concentrations were measured at different points in the study, and as a result, different platforms were used based on effectiveness. A form was used. Several serum measurements were performed using Roche Elecsys 2010. In NALISA (Roche Diagnostics, Laval, Canada) The test was performed using an electrochemiluminescence immunoassay. Additional serum measurements were performed using miniVIDAS immunoassay. The assay was performed using a No assay (bioMerieux, Durham NC, USA). Blood If Qing is unavailable, B·R·A·H·M·S PCT-sensitive KRYPTOR (The Using RMO Fisher Scientific (Portage MI, USA) Immunofluorescence in plasma-EDTA is used in Phadia Immunology Ref. Measurements were performed by the Erance Laboratory. Several pairs of serum and Repeated tests were conducted on plasma samples, and equivalence in concentration was revealed. Therefore, All procalcitonin measurements were treated equally, regardless of the testing platform.

[0155] Microarray fabrication At the initial clinical findings, the patient was registered and a sample was collected for analysis. The determination was made as described above. After conducting the procedure, 317 subjects with a clear clinical phenotype were subjected to gene expression analysis. I chose this for the following reasons: Total RNA was collected using the PAXgene blood RNA kit (Qiagen, Vale Extracted from human blood using ncia (CA) according to the manufacturer's protocol. RN The quantity and quality of A were measured using a Nanodrop spectrometer (Thermo Scientometer). ific, Waltham, MA) and Agilent 2100 Bioanalyzer The evaluation was performed using Agilent, Santa Clara, CA. RMA normalization was performed. Hybridization and data collection were carried out using Expression. Analysis (Durham, NC) shows that GeneChip human genome U1 Using the 33A 2.0 array (Affymetrix, Santa Clara, CA) Then, we carried it out according to the Affymetrix technical manual.

[0156] statistical analysis Transition of 317 subjects (273 patients with illness 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 were used to obtain substantial data. The difference was revealed. These were analyzed using a Bayesian fixed-effects model in relation to the probe. First, the average batch effect was estimated and corrected by removing it. Next, the inventors They used seven duplicate samples to fit a robust linear regression model to the Huber loss function. The expression levels were adjusted using the resulting mixture.

[0157] Sparse classification methods such as sparse logistic regression have a risk of overfitting. While reducing the value, classification and variable selection are performed simultaneously. 21 Therefore, a univariate test is not possible. Therefore, separate gene selection strategies such as sparse factor models are unnecessary. The sparse logistic regression model was batch-corrected and the probe with the maximum variance was selected from 40 We used % to fit each of the binary tasks separately. 22 Specifically, this The authors developed a Lasso-regularized generalized linear model with binomial likelihood, along with nested cross-validation. The regularization parameters were selected using a model. The code is from the Glmnet toolbox. Written in Matlab using .x. This is a bacterial ARI classifier, viral ARI This generated an I classifier and a non-infectious disease classifier. Each binary classifier has class members. Probability (for example, in the case of a bacterial ARI classifier, whether it is viral or non-infectious) Under the condition of estimating the probability of bacterial infection, the inventors have determined the maximum membership probability The rate follows a one-to-many scheme to assign class labels, thereby assigning the three classifiers ( It can be integrated into a single decision model (named ARI classifier). 21 . classification ability measurement The standard criteria are the area under the receiver operating characteristic curve (AUC) for the binomial result and the ternary result. Includes confusion matrix of the same 23 .

[0158] verification The ARI classifier was validated using one-out cross-validation within the same population from which it was derived. The publicly available human gene expression dataset (GSE626) from 328 individuals was used. 9, GSE42026, GSE40396, GSE20346, and GSE42834 Independent external validation was performed using ). The datasets were determined to be at least two clinical trials. Selected when the bed group includes bacterial ARI, viral ARI, or non-infectious diseases. To align probes across different microarray platforms, each ARI classifier probes are converted into gene symbols, and these are used to target the corresponding micro The array probe was identified.

[0159] result Bacterial ARI classifiers, viral ARI classifiers, and non-infectious disease classifiers In creating classifiers based on host gene expression to distinguish between clinical states, all The relevant clinical phenotype should be expressed during the model training process. This should be assigned specificity. This allows the model to be applied to these included clinical groups, but model training It will not be possible to apply this to clinical phenotypes that did not exist. 15 Targets for ARI diagnosis The group includes not only patients with viral and bacterial etiologies, but also other possibilities. It must also be distinguished from patients who do not have bacterial or viral ARIs. From a general perspective, healthy individuals have served as a control group that has not been infected. However, However, this is how patients with non-communicable diseases that may present with similar clinical symptoms are classified. The fact that this will happen has not been taken into consideration, which is a potential source of diagnostic errors. To the best of the inventors' knowledge So, a classifier based on ARI gene expression, if the disease is not present in its derivation... None of the data include controls. Therefore, the inventors have identified community-acquired viral ARIs (n= 115) First, patients with bacterial ARI (n=70) or non-infectious diseases (n=88) A large, heterogeneous group of patients at the time of clinical findings was registered (Table 8). The inventors also To define the most appropriate control population for ARI classifier development, a healthy adult control cohort This includes (n=44).

[0160] The inventors first determined that the gene expression classifier derived using healthy individuals as a control was non We determined whether it was possible to accurately classify patients with infectious diseases. Bacterial ARI, viral Using array data from patients with seroconjunctival arrhythmia and healthy controls, these conditions We created a gene expression classifier for this. Using one-miss cross-validation, we achieved an overall accuracy of 90%. And, bacterial ARI (AUC 0.96), viral ARI (AUC 0.95), Highly accurate differentiation between healthy and unhealthy subjects (AUC 1.0) was demonstrated (Figure 7). However, the classifier is ill-uninfected (ill-uninfected). When applied to patients with tediatric infections, 48 / 88 are bacterial, 35 / 88 are viral, and 5 out of 88 were identified as healthy. This means that healthy individuals are in the biomarker discovery process. This highlighted the fact that it cannot be used as a substitute for patients with non-communicable diseases.

[0161] Consequently, the inventors used controls with non-infectious diseases rather than healthy controls to study ARI The classifier was re-derived. Specifically, using array data from these three groups, Three genes related to the host response to bacterial ARI, viral ARI, and non-infectious diseases. Expression classifiers were created (Figure 5). Specifically, the bacterial ARI classifier was created for viral ARI Or the task of clearly identifying individuals with bacterial ARI for either non-communicable diseases. It was imposed. The viral ARI classifier is used for bacterial ARI or non-infectious diseases. The task was to clearly identify individuals with anemic ARI. The non-infectious disease classifier is all This means clearly identifying all non-communicable diseases that require sufficient representation of such cases. It was not created in the diagram.

[0162] Rather, it was created as an alternative category, and therefore, bacterial ARIs are also viral Patients who do not have serotonin-related ARIs may be assigned. Furthermore, the inventors believe that healthy people can also be assigned. It is unlikely that it is the target of such classification tasks, therefore, it is unlikely that it has such a disease. We hypothesized that uninfected 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 Foreth These gene expression classifiers were created using six comprehensive strategies. They all worked the same way, but sparse logistic regression used the fewest classifiers It required genes and performed slightly better than other strategies (data not yet presented). (Demonstration) The inventors also have a strategy for creating three distinct binary classifiers for a given number of pairs This was compared to a single multinomial classifier that simultaneously assigns elephants to one of three clinical categories. The latter approach required more genes and was less accurate. The inventors applied a sparse logistic regression model to obtain 71, 33, and Each contains 26 probe signatures, including a bacterial ARI classifier and a viral A RI classifiers and non-infectious disease classifiers were defined. Probe and classifier weights are shown in Table 9. It is shown.

[0164] Clinical decision-making is rarely binary and requires the simultaneous distinction of multiple diagnostic possibilities. The inventors assigned probabilities to bacterial ARI, viral ARI, and non-infectious diseases. To do this, we use cross-validation with one exception removed, and together we define all three ARI classifiers. The classifier was applied (Figure 6). These conditions are not mutually exclusive. For example, bacterial AR The presence of I does not prevent simultaneous viral ARI or non-infectious diseases. Furthermore, the assigned probability is that the patient's gene expression response matches the canonical signature of that condition. Represents the degree to which it matches the target. Each signature functions intentionally, independently of the others. Therefore, the probabilities are not intended to be summed into one. To simplify the classification, most High prediction probability determined the class assignment. The overall classification accuracy was 87% (238 / 273). (However, it matched the determined phenotype.)

[0165] Bacterial ARI was identified in 58 / 70 (83%) of patients who did not have a bacterial infection. 179 / 191 (94%) were excluded. Viral ARIs accounted for 90% (104 / 115). It was identified and excluded in 92% (145 / 158) of cases using a non-infectious disease classifier. Infection was ruled out in 86% (76 / 88) of cases. Prevalence was determined by infection type, patient characteristics, and Considering that it can vary for a number of reasons, including location, all three classifiers are explicit Sensitivity analysis was performed on the sexual accuracy and negative predictive value (Figure 8). Bacterial classification and viral classification For both categories, the predictive value is high, including the broad prevalence typically found in ARI. The rate remained high across the estimation range.

[0166] To determine whether there is any age-related effect, the inventors have created a classification scheme. It was included as a variable in this. This resulted in two additional positive classifications, however This is likely because young people constituted a large proportion of the viral ARI cohort. However, the inventors mistakenly identified correctly classified subjects by age. No statistically significant differences were observed among the classified subjects (Wilcoxon rank sum p=0.1). 7).

[0167] The inventors have demonstrated this performance using a widely used biomarker specific to bacterial infections. It was compared with a certain procalcitonin. The procalcitonin concentration was compared with 23 samples that were available. We made decisions for eight subjects and compared the ARI classifier performance for this subgroup. Procalcitonin concentrations higher than 0.25 μg / L indicate that the patient has bacterial ARI. On the other hand, values ​​below 0.25 μg / L indicate that the patient has either viral ARI or a non-infectious disease. It was assigned to either nonbacterial or non-bacterial. Procalcitonin was classified using the ARI classifier. Compared to 204 / 238 (86%), 186 out of 238 patients (78%) were correct. They were classified differently (p=0.03). However, the accuracy of those two strategies was poor. This varied depending on the classification task. For example, distinguishing viral ARIs from bacterial ARIs. In this respect, the performance was similar. Procalcitonin was 140 / for the ARI classifier. Compared to 155, 136 / 155 (AUC 0.89) was correctly classified (Yates' supplement). Using McNemar's test for positive results, the p-value was 0.65. However, the ARI classifier This distinguishes bacterial ARIs from non-infectious diseases [105 / 124 vs 79 / 124 (A [UC 0.72); p-value < 0.001], and bacterial ARI compared to viral and non-infectious. Distinguishing from all other etiologies, including infectious etiologies [215 / 238 vs 186 / 238 (A In a UC of 0.82 (p-value = 0.02), it was significantly superior to procalcitonin. .

[0168] The inventors then developed the ARI classifier based on three published genetic classifications for bacterial infection versus viral infection. The results were compared with sub-expression classifiers (each derived without including uninfected disease controls). These are 35 probes derived from children with influenza or bacterial sepsis. The classifier (Ramilo) 11 From children with a febrile viral illness or bacterial infection The derived 33-probe classifier (Hu) 14 ; and community-acquired pneumonia or influenza Classifiers (Parnell) of 29 probes derived from adult ICU patients 12 Includes The inventors have developed a classifier made using only patients with viral or bacterial infections. However, this was applied to a clinically relevant population, including patients who had the disease but were not infected. In that case, I hypothesized that it wouldn't work well. Specifically, both bacterial and viral infections Regarding individuals that do not possess this classification, the previously published classifiers classify those individuals into a third, separate category. It would not be possible to assign it precisely to Lee. Therefore, the inventors have derived the The inventors applied the classifiers, along with the published classifiers, to a cohort of 273 patients. Distinguishing between sexual ARIs, viral ARIs, and non-infectious diseases is based on the derived ARI classification. It was superior in terms of similarity (McNemar's test with Yates correction, Ramilo) For p=0.002; for Parnell, p=0.0001; and 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 a cohort that represents a particular intended use population. It highlights the essential nature of the matter. 15 .

[0169] Classification of discrepancies To better understand the performance of the ARI classifier, the inventors examined 35 cases of mismatch. Each case was re-examined. Nine of the identified bacterial infections were classified as viral, and three were non-infectious. It was classified as a disease. Four viral infections were classified as bacterial, and seven were classified as non-infectious. Eight non-infectious cases were classified as bacterial, and four were classified as viral. The researchers did not observe a consistent pattern among the discrepant cases, but there were some notable cases that were typical. It was included in non-target bacterial infections. One person had M. pneumoniae based on serological conversion. One of the patients, and one of the three patients with Legionnaires' disease, had viral ARI and Classified. Six patients with non-infectious diseases due to autoimmune or inflammatory diseases. Of the individuals diagnosed with Still's disease, only one was classified as having a bacterial infection. More specifically incorporated herein, Tsalik et al., (2016) Sci Tran See also eTable 3 in sl Med 9(322):1-9.

[0170] External Verification Creating classifiers from high-dimensional gene expression data can lead to overfitting. Therefore, the inventors have found five available datasets (GSE6269, GS (E42026, GSE40396, GSE20346, and GSE42834) odor Using gene expression data from 328 individuals represented by [the model name], an ARI classifier was created in silico. This was examined. These differ in terms of age, geographical distribution, and disease severity. The ARI classifier was selected from those that included at least two related clinical groups (Table 7). When applied to four datasets containing bacterial and viral ARIs, AUC The range was 0.90 to 0.99. Finally, GSE42834 caused bacterial pneumonia (n= 19) Includes 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 5 subjects with bacterial pneumonia before and after treatment. All cases showed treatment-dependent resolution of bacterial infection. Tsalik et al., (2016) Sci Transl Med 9(322):1 See also eFigures 3-8 in -9.

[0171] Biological pathways The sparse logistic regression model that produced the classifier is not useful if there is no additional diagnostic value. This disadvantages the selection of genes from a predetermined pathway. Consequently, normal gene enrichment Pathway analysis is not appropriate to perform. Furthermore, such conventional gene enrichment A comment analysis is provided. 9,12,14,28,29 Instead, all classifier inheritance A literature review was conducted on the following (Table 10): bacterial classifiers, viral classifiers, and non- The overlaps among infectious disease classifiers are shown in Figure 9.

[0172] Viral classifiers are involved in interferon response, T cell signaling, and RNA processing. This included known antiviral response categories such as Sing. The viral classifiers were related to nuclear transport. K is selected by the virus for the transport of viral proteins and genomes. It best represents RNA processing pathways such as PNB1. 26,27 Its downward This suggests that it may play an antiviral role in the host response.

[0173] Bacterial taxonomies encompass the broadest range of cellular processes, particularly cell cycle control, cell proliferation, and differentiation. It included. Bacterial taxonomies are important genes in T cell, B cell, and NK cell signaling. It contained genes. Specific to the bacterial taxonomy was oxidative stress, consistent with sepsis-related metabolic perturbations. , as well as genes involved in fatty acid and amino acid metabolism. 28 .

[0174] Overview of Clinical Applicability The inventors believe that changes in host gene expression are highly specific to the causative pathogen class, and It was determined that it could be used to identify the common etiologies of respiratory diseases. This is inappropriate. Novel diagnostic platform to curb excessive antibiotic use and the emergence of antibiotic resistance Develop gene expression classifiers as a form and create opportunities for their use. Sparse logistic Using hack regression, the inventors have identified a bacterial etiology in patients with acute respiratory symptoms and We developed a host gene expression profile that accurately distinguishes viral etiologies (external validation AU). C 0.90~0.99). Deriving ARI classifiers using non-infectious disease control groups is a wide range. It yielded a high negative predictive value across a wide range of prevalence estimates.

[0175] Respiratory tract infections were more common than any other cause of illness in 2011, accounting for 3.2 million cases worldwide. This resulted in human death and the loss of 164 million disability-adjusted life years. 1,2 Most of the cases Despite the fact that the cause is viral, in the United States, there are cases of acute respiratory infections (ARIs). 73% of outpatients were prescribed antibiotics, and of all antibiotics prescribed in this setting... 41% 3,4 Even when the viral pathogen has been microbiologically identified, This does not rule out the possibility of simultaneous bacterial infection, and the antibacterial agent is "just in case." This leads to prescriptions. This empiricism gives rise to antimicrobial resistance. 5,6 That is a matter of national security. It is recognized as a priority. 7 The promising measurement method provided in this embodiment is for the treatment A device that optimizes and reduces the emergence of antibiotic resistance, providing clinically immediate results. To bring about a meeting.

[0176] Several studies have made noteworthy efforts to develop host response diagnostics for ARI. This became an initiative. This is a respiratory virus. 8,10~12,14 In the ICU population bacterial etiology 12,30 , and tuberculosis 31~33 Includes responses to these. Typically, these This defines the host response profile compared to a healthy state and provides valuable insights into the host ecosystem. provide 16,34,35 However, these gene lists are relevant to diagnostic application. It is suboptimal, but that is because the gene expression profile, which is a component of the diagnosis, is tested Because it does not represent the group to which it would apply. 15 Healthy individuals are prone to acute respiratory illnesses. They did not show any signs of disease, and therefore they were excluded from the host response diagnostic developments reported herein. To be removed.

[0177] Including patients with bacterial and viral infections, the distinction between these two conditions While this allows for differentiation, it does not address how to classify non-communicable diseases. Patients share symptoms. Including this phenotype is important to indicate possible infectious and non-infectious etiologies. In other words, symptoms may not provide physicians with a high degree of diagnostic certainty. ARI symptoms A book that uniquely understands the need to include the three most likely states regarding this. The approach is to apply such trials to undifferentiated clinical populations where they are most desperate. It is possible.

[0178] The few discrepancies in classification that occurred were either due to errors in classification or errors in clinical phenotyping. This may have resulted from the limitations of current microbiological diagnostics. Errors in clinical phenotypic testing are due to the limitations of current microbiological diagnostics. This can result from the inability to identify the causative pathogen. Alternatively, it can be due to some non-infectious factors. The disease process may actually be infection-related, due to mechanisms that have not yet been discovered. The examples were not clearly explained by unifying variables such as pathogen type, syndrome, or patient characteristics. Therefore, the gene expression classifiers presented herein are patient-specific variables (e.g., Treatment, comorbidities, duration of illness), test-specific variables (e.g., sample preparation, assay conditions, RN) A) Quality and yield), or other factors including variables that have not yet been identified. It could potentially be affected.

[0179] [Example 2] Patients with co-infections defined by the identification of bacterial and viral pathogens Classification ability In addition to determining that age does not significantly affect classification accuracy, the inventors The study assessed whether disease severity or the etiology of SIRS influenced the classification. (Virus) Patients with sexual ARI have less disease, as evidenced by lower hospitalization rates. There was a tendency not to develop the disease. In various cohorts, hospitalization was used as a marker of disease severity. The study used this method and evaluated its impact on classification ability. This study revealed no difference. (Fisher direct test 1 p-value). In addition, the SIRS control cohort showed respiratory etiology. The present inventors have included both subjects with respiratory SIRs and subjects with non-respiratory etiologies. Evaluate whether there is a difference in classification between subjects with S and subjects with non-respiratory SIRS. Therefore, it was determined that they were not different (Fisher's direct test p-value of 0.1305).

[0180] Some patients with ARI often have co-infections, which are bacterial pathogens and viruses. Both sexual pathogens are identified. However, how the host responds in such a situation It is unclear whether this is a response. The disease manifests itself at different points in the patient's clinical course. It can be manipulated by bacteria, viruses, both, or neither. Therefore, this The inventors have found that bacterial ARI classifiers and viral ARI classifiers are similar in that they can distinguish between bacteria and viruses. We determined how it functions in populations including time-borne infections. GSE60244 is a bacterium. Pneumonia (n=22), viral respiratory tract infections (n=71), and bacterial / viral infections This included time-identification (n=25). The simultaneous identification group was related to the likelihood of bacterial or viral disease. Without further subcategorization, the existence of both bacterial and viral pathogens Defined by [the present invention]. The inventors of GSE60244 have a bacterial or viral infection. The classifier was trained on subjects in the specified area, and then validated on subjects with simultaneous identification capabilities. (Figure 10). The host response was considered positive, exceeding a probability threshold of 0.5. The inventors, All four possible categories were observed. Six out of 25 subjects tested positive for bacterial stool. They had a gum-like texture, 14 / 25 had a viral response, and 3 / 25 were positive for bacterial and viral infections. It possessed a Rustic signature, and neither of the 2 / 25 showed it.

[0181] A critical clinical decision that physicians face is whether or not to prescribe antibacterial agents. The discontinuation strategy is based solely on the probability of bacterial ARI according to the results from the bacterial ARI classifier. It can be focused on. However, information about other possibilities, whether viral or non-infectious, is needed. Providing information is valuable. For example, if the probability of an alternative diagnosis is high, the probability of bacterial ARI is high. We can have greater confidence in withholding antibacterial agents in patients with low levels of risk. Furthermore, a complete diagnostic report can identify symptomatic diseases that a single classifier might miss. This is possible. The inventors have verified this in a population including the simultaneous identification of bacteria and viruses. This was observed. These patients are more commonly referred to as "comorbidly infected." In order for infection to occur, there must be a maladaptive interaction between the pathogen, the host, and the two. They must be there. Simply identifying bacterial and viral pathogens is not enough to distinguish co-infections. This should not be implied. The inventors have found evidence of simultaneous identification of bacteria / viruses in their tests. Although the true infection status of the 25 subjects cannot be determined, the host response classifier is: This suggests the existence of multiple host response states. Figure 10 is a diagram that provides information about the infection state. It can be used by clinicians to diagnose the pathogenesis of ARI.

[0182] References JPEG0007849855000025.jpg225147 JPEG0007849855000026.jpg231151 JPEG0007849855000027.jpg145149

[0183] JPEG0007849855000028.jpg93162

[0184] JPEG0007849855000029.jpg89162

[0185] JPEG0007849855000030.jpg234147 JPEG0007849855000031.jpg167149

[0186] JPEG0007849855000032.jpg235131 JPEG0007849855000033.jpg234143 JPEG0007849855000034.jpg235147JPEG0007849855000035.jpg235147 JPEG0007849855000036.jpg235145JPEG0007849855000037.jpg23554

[0187] JPEG0007849855000038.jpg194155

[0188] [Example 3] Bacterial / viral / SIRS assays designed for the TLDA platform The inventors have developed a 384-well TaqMan low-density array (TLDA, Applied Custom multi-analyte quantitative real-time analysis on the Biosystems platform. We are trying to develop an immunoPCR (RT-PCR) assay. TLDA cards are available. Along with multiple endogenous control RNA targets (primer / probe sets) for normalization, In each well, one or more T genes specific to the mRNA transcript of the classifier are found. It is prepared using the aqMan primer / probe set. The entire RNA is reverse transcribed into cDNA, loaded into the master well, and then processed in a microfluidic system. The solution is distributed to each assay well by centrifugation through the channel. TaqMan hydration is performed. The resolution probe is dual during hybridization with complementary target sequences in each amplification round. The labeled probe is cut, resulting in the generation of a fluorescent signal from 5' to 3'. It depends on exonuclease activity. In this manner, it accumulates in "real time". Quantitative detection of PCR products is possible. During exponential amplification and detection, a fluorescence signal is detected. The number of PCR cycles at which the detection threshold is exceeded is the number of commercially available RT-PCR instruments. Threshold cycle (C) when determined by software t ) or quantification cycle (C q ) Therefore, to quantify gene expression, C for target RNA is used. t However, intrinsic normalization R C of NA (or the geometric mean of multiple normalized RNAs) t It is subtracted from the sample, and thereafter, δC for each RNA target within t A value is given, which is the input sample RNA or cDN. This shows the relative expression of the target RNA normalized for variations in the quantity or quality of A.

[0189] The data on the quantified gene signatures was then processed using a computer. The following probit classifier (Equation 1) is processed and reproduced here. Signature The normalized gene expression level of each gene is used as an explanatory variable or independent variable in the classifier. These are features, and in this example, the general type of classifier is the following probit regression equation is: P(having condition)=Φ(β1X1+β2X2+…+β d X d )(equation 1) In the formula, the condition is bacterial ARI, viral ARI, or non-infectious disease; Φ(.) is a probit link function; {β1, β2, ..., β d} is a coefficient obtained during training. ri;{X1,X2,…,X d} is the normalized gene expression value of the signature; d is the signature This is the size of the gene (number of genes). The coefficient values ​​for each explanatory variable are given by the probit regression model. Techniques used to measure the expression of genes or subsets of genes used in Dell It is specific to the Nology platform. The computer program scores or confirms Calculate the rate and compare the score to the threshold. Calculate the sensitivity, specificity, and overall precision of each classifier. The degree can be determined by changing the threshold for classification using the receiver operating characteristic (ROC) curve. It will be optimized.

[0190] Signatures from the Affymetrix platform (Affy signatures) Genes for TLDA platforms based on signatures from other sources A preliminary list is provided in Table 1A below. TLDA profiles for each classifier. The appropriate weight for the foam is then determined as described in Example 4 below. It was decided.

[0191] JPEG0007849855000039.jpg232151JPEG0007849855000040.jpg240148JPEG0007849855000041.jpg221148

[0192] [Example 4] Genetic data measured by RT-qPCR, as intended in the TLDA platform. Bacterial / viral / SIRS classification using sub-expression The three signature genes that make up the host response-ARI (HR-ARI) test are Th Custom made by ermoFisher Scientific (Waltham, MA) These gene signatures have been transferred to TaqMan® low-density array cards. Expression was controlled using a 384-well TaqMan low-density array (TLDA; Thermo-Fishe r) Custom multianalyte quantitative real-time PCR (RT-q) on the platform Measurements were taken using a PCR assay. TLDA cards were normalized for RNA loading. Multiple intrinsic properties used to regulate inter-plate variability Control RNA targets (TRAP1, PPIB, GAPDH, FPGAS, DECR1, and 1 Along with 8S), one or more TaqMan primer / probe sets per well. Designed and manufactured using ARI Signature, each primer / probe set is made with ARI Signature. This represents a specific RNA transcript in the sample for normalization. Then, we select two reference genes (out of the five available) that have the smallest coefficient of variation, and 3 Primer / probe sets with more than 3% missing values ​​(below the limit of quantification) were discarded. (If any) the remaining missing values ​​are 1 + max(C) q )(C q This is about RT-qPCR Set to a quantitative cycle. Then, normalize the expression value using any predetermined primer / Observed C for the probe set q Calculated as the average of the selected references after subtracting the values. 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 primers / probes, 144 primer / probe sets are 3 AR Affymetrix (microarray) described 132 terms before the I gene signature ) Gene targets representing probes (i.e., bacterial gene expression signatures, viral gene expression The current signature and the genes in the non-infectious gene expression signature were measured, and six The probe set is for reference genes, and the inventors have added previously discovered probes. A set of 24 probes from the panviral gene signature was assayed. No. 8,821,876; Zaas et al., Cell Host Microbe (2009 See 6(3):207-217. In addition, several ply for “replacement” genes. MARS / Probe Set (These gene expressions are derived from the Affymetrix signature.) We added several genes (correlated with the expression of some genes) for training. Although they are replaced, this is done using TLDA probes, and these genes This was because the RT-qPCR assay for the offspring did not work properly.

[0194] For each sample, the total RNA was collected in a PAXgene blood RNA tube (PreAnalyt ix) Purified from the Superscript VILO cDNA synthesis kit (Ther Using mo-Fisher, the cDNA is converted according to the manufacturer's recommended protocol. Reverse transcription was performed. A standard amount of cDNA for each sample was loaded into each masterwell, and The solution is then distributed to each TaqMan assay well via centrifugation through a chlorofluid channel. TaqMan hydrolysis probes hybridize with complementary target sequences in each amplification round. The double-labeled probe is cleaved during dilation, resulting in the generation of a fluorescence signal. It depends on the exonuclease activity from 5' to 3'. Quantitative detection of fluorescence is " The accumulated PCR product is shown in "altime". During exponential amplification and detection, fluorescence is detected. The number of PCR cycles at which the signal exceeds the detection threshold is the market value for RT-PCR instruments. Threshold cycle (C) when determined by the sales software t ) or quantification cycle ( C q)

[0195] Sample / cohort selection: Under an IRB-approved protocol, the inventors have developed a method for treating patients with acute respiratory illnesses in the emergency department. Patients who were examined were registered (see Table 11 below). Patients in this cohort were referred to by this reference. The following is incorporated into the specification: Tsalik et al., (2016) Sci Transl M This is a subset of patients reported in Table 1 of ed 9(322):1-9. Retrospective clinical assessment of clinical and other test data indicates that bacterial ARI, viral One of three assignments is derived, which is either a septic ARI or a non-infectious disease.

[0196] JPEG0007849855000042.jpg94154

[0197] Data analysis methods: During the data preprocessing stage, the inventors have obtained the smallest coefficient of variation across the sample and plate. Select at least two subsets of reference gene targets (out of the five available). The inventors identified a target with more than 33% missing values ​​(17 targets below the limit of quantification). These values ​​account for a large proportion in any particular class, for example, bacterial ARI. Only in the case of [this] was it discarded. Next, the inventors calculated the remaining missing values ​​as 1 + max(C) q Set to ) Then, the expression values ​​for all targets were normalized using the previously selected reference combination. The inventors have determined that the normalized expression value is the C of any predetermined target. q Selected reference with value subtracted ( It was calculated as the average of DECR1 and PPIB.

[0198] Once the data is normalized, the inventors can use sparse logistic regression. We will proceed to build a classification model by fitting Dell to that data. (Friedman et al. (2010) J. Stat. Softw. 33, 1- 22). This model calculates the probability that an object belongs to a specific class as a weighted sum of normalized gene targets. We estimate that p(object belongs to class) = σ(w1x 1 + ... + w p x p ) is written as such, and in the formula, σ is the logistic function, and w1, …, w p x1, …,x are the classification weights estimated during the fitting procedure. p This is the normalized expression value. This represents p gene targets containing [the specified element].

[0199] Similar to array-based classifiers, the inventors have identified (1) bacterial ARI vs. viral A (1) RI and non-infectious diseases; (2) Viral ARI vs. bacterial ARI and non-infectious diseases; Furthermore, (3) construct three binary classifications for non-infectious diseases versus bacterial and viral ARIs. After fitting the three classifiers, the inventors determined that p (bacterial ARI), Obtain estimates for p (viral ARI) and p (non-infectious disease). The thresholds for each are selected from the receiver operating characteristic (ROC) curve using a symmetric cost function. Selected (expected sensitivity and specificity are approximately equal) (Fawcett (2006)) Pattern Recognition Lett 27:861-874). As a result, Elephant is p(bacterial ARI)>t b (t b (This is the threshold for bacterial ARI classifiers) It is predicted to be a bacterial ARI. Similarly, the inventors predict viral ARI classifiers and non-infectious ARIs. Thresholds for sexually transmitted disease classifiers, respectively, tv and t n Select the option. If necessary, The conditions that are most likely to be met, that is, the conditions with the highest probability (specifically, the inventors is argmax{p(bacterial ARI), p(viral ARI), p(non-infectious disease)} By selecting (write as), you can perform integrated forecasting.

[0200] result: The first TLDA platform to incorporate genome classifiers discovered by microarrays During the conversion, the inventors used 32 samples that had also been assayed by microarrays. This group is based on the TLDAs of the gene transcripts that make up the ARI classifier. RT-qPCR measurement, which was previously used, is now available for measurement based on microarrays of gene transcripts. The results were reproduced, and therefore, the patient had either bacterial or viral ARI. It is a valid methodology for classifying individuals as having a non-infectious disease or as having a non-infectious disease. It plays a role in recognition. The inventors have developed a TLDA platform and a microarray platform. From 32 samples tested in both forms, their corresponding classifiers were used. When evaluated, there was an 84.4% agreement, meaning that 27 out of 32 subjects agreed. This applies to both microarray-based and TLDA-based classification models. This indicates that it showed a combined prediction.

[0201] After demonstrating the agreement between classification based on microarrays and classification based on TLDA, The inventors have a clinical assessment of the ARI state, but the previously characterized gene expression pattern An additional 63 samples from patients who did not show symptoms were tested using a classification based on TLDA. Therefore, a total of 95 samples were evaluated using a classification test based on TLDA. This dataset from 95 samples allows for TLDA-based RT-qPCR testing. The rat form classifies new patients based solely on clinical judgment as a reference standard. It became possible to use and evaluate this. In this experiment, the inventors found that 77 out of 95 We observed an overall accuracy of 81.1% for correctly classified samples. More specifically, The models were 80% (24 out of 30 correct) and 77.4% (31 24 out of 34 were correct, and 85.3% (29 out of 34 were correct) were bacterial. ARI accuracy, viral ARI accuracy, and non-infectious disease accuracy were obtained for each classifier. Regarding performance, the inventors have demonstrated that bacterial ARI classifiers, viral ARI classifiers, and non For each infectious disease classifier, the ROC values ​​are 0.92, 0.86, and 0.91. The area under the line was observed. The inventors used a validation dataset for each of the classifiers. Without counting, the inventors still found the classification ability (accuracy and area under the ROC curve) to be insufficient. If partial estimates are desired, the inventors report a performance measurement criterion method that has been cross-validated with one sample removed. They are doing it.

[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 174 genes. Please note that instead of a specific target, 151 gene targets are listed. As described, the reference gene was removed during the pretreatment stage, and there were 17 missing values. This is because the individual targets were also removed in the same manner. These 17 targets were also removed during the pre-processing stage. Removed.

[0203] If the panviral signature gene is removed, the AUC, precision, and match values ​​will be A slight performance decrease of at most 5% is observed across the percentage range.

[0204] summary: The composite host response ARI classifier includes bacterial ARIs and non-infectious diseases, as opposed to viral ARIs. Gene expression signatures and mathematical classification frameworks are used to diagnose bacterial ARIs. It consists of a mathematical classifier that determines whether the subject is bacterial ARI, viral ARI, or non- It provides three separate probabilities of having an infectious disease. In each case, the cutoff is... Alternatively, a threshold can be identified, and a value above that threshold is determined to indicate that the patient has that condition. Furthermore, the threshold can be modified to change the sensitivity and specificity of the test.

[0205] These gene expression level measurements can occur on various technological platforms. Here, the inventors use a TLDA-based RT-qPCR platform. The measurements of these signatures are described. Furthermore, the mathematical framework for determining the ARI etiology probability is described. The teamwork involves platform-specific training to adapt the transcript measurement method. platform-specific weights, w1, …,w p Platt (by establishing) Adapt to the form. Similar direct methodologies use gene signatures to transform other gene expression platforms. It can be converted into a template, and then the associated classifiers can be trained. This example also includes a TLDA-based classification of the etiology of ARI and a microarray-based classification. This demonstrates excellent agreement with the classification. Finally, the inventors have found that patients with acute respiratory disease A TLDA-based RT-qPCR platform for diagnosing new patients This demonstrates the use of related mathematical classifiers.

[0206] JPEG0007849855000043.jpg236132 JPEG0007849855000044.jpg234146 JPEG0007849855000045.jpg235139JPEG0007849855000046.jpg235149JPEG0007849855000047.jpg235146JPEG0007849855000048.jpg235142 JPEG0007849855000049.jpg234143JPEG0007849855000050.jpg235141JPEG0007849855000051.jpg235146 JPEG0007849855000052.jpg235105

[0207] Any patents or publications referenced in this specification represent a level of skill in the art. The patents and publications are designed to appear as if each individual publication is incorporated by reference. As shown concretely and individually, incorporated herein by reference to the same extent In case of any conflict, this specification, including its definitions, shall prevail.

[0208] This invention aims to perform the subject matter and the purposes and benefits mentioned, in addition to those specific thereto. Those skilled in the art will readily recognize that it is necessary to adapt the system sufficiently to obtain the following. The present disclosure represents preferred embodiments, is illustrative, and does not extend to the scope of the present invention. It is not intended as a limitation. Its variations and other uses may be conceivable to those skilled in the art. They are encompassed within the spirit of the invention as defined by the claims. It is included.

Claims

1. A method for determining the likely etiology of an acute respiratory disease, selected from bacterial, viral, and / or non-infectious causes, in a person suffering from or at risk of suffering from it, (a) A step of preparing a peripheral blood sample obtained from the subject, (b) A step of measuring the gene expression levels of a defined set of genes in the peripheral blood sample on the platform. (c) A step of normalizing the gene expression level to obtain a normalized gene expression value. (d) Inputting the normalized gene expression values ​​into one or more acute respiratory disease classifiers selected from bacterial acute respiratory infection (ARI) classifiers, viral ARI classifiers, and non-infectious disease classifiers, wherein the classifier includes defined weighting values ​​for each of the genes in the defined set of genes for the platform, The bacterial ARI classifier includes expression levels of at least five genes with weighted values ​​listed as part of the bacterial classifiers in Table 1, Table 9, and / or Table 12. The viral ARI classifier includes at least 10 expression levels of genes with weighted values ​​listed as part of the viral classifiers in Table 1, Table 9, and / or Table 12. The steps include: the non-infectious disease classifier includes the expression levels of at least five genes with weighted values ​​listed as part of the non-infectious disease classifiers in Table 1, Table 9, and / or Table 12; and (e) A step of calculating the probability of pathogenicity for one or more bacterial ARIs, viral ARIs, and non-infectious diseases based on the normalized gene expression values ​​and the classifiers. A method comprising, by which determining the likelihood that the acute respiratory disease in the subject is of bacterial, viral, non-infectious origin, or some combination thereof.

2. (f) The step of comparing the probability with a range of predefined thresholds, cutoff values, or values ​​indicating the likelihood of infection. The method according to claim 1, further comprising:

3. The method according to claim 1 or 2, wherein the subject is suffering from symptoms of an acute respiratory disease.

4. The method according to any one of claims 1 to 3, wherein the subject is suspected to have a bacterial or viral infection.

5. The method according to any one of claims 1 to 4, further comprising the step of repeating steps (d) and (e) using only a viral classifier and / or a non-infectious classifier if the peripheral blood sample does not indicate a bacterial ARI, to determine whether the acute respiratory disease in the subject is of viral origin, non-infectious origin, or a combination thereof.

6. The method according to any one of claims 1 to 4, further comprising the step of repeating steps (d) and (e) using only bacterial classifiers and / or non-infectious classifiers if the peripheral blood sample does not indicate a likelihood of viral ARI, to determine whether the acute respiratory disease in the subject is of bacterial origin, non-infectious origin, or a combination thereof.

7. The method according to any one of claims 1 to 4, further comprising the step of repeating steps (d) and (e) using only bacterial and / or viral classifiers if the peripheral blood sample does not indicate a likelihood of non-infectious disease, to determine whether the acute respiratory disease in the subject is of bacterial origin, viral origin, or a combination thereof.

8. The method according to any one of claims 1 to 7, further comprising the step of preparing a report that assigns a score indicating the probability of the etiology of the acute respiratory disease to the subject.

9. The method according to any one of claims 1 to 8, wherein the defined gene set comprises 30 to 200 genes.

10. The method according to any one of claims 1 to 9, wherein the defined gene set comprises 30 to 200 genes listed in Table 1, Table 9, and / or Table 12.

11. The method according to any one of claims 1 to 10, comprising the step of monitoring the response of the subject to a vaccine or drug.

12. The method according to claim 11, wherein the drug is an antibacterial agent or an antiviral agent.

Images