WO2018154075A1 - Methods for classifying subjects exposed to viral infection - Google Patents

Abstract

The present invention provides methods for classifying subjects exposed to viral infection, in particular influenza. The methods may be used to classify subjects before they show symptoms of infection. The methods may also be used to classify subjects who are infected but are asymptomatic. The present invention also provides an apparatus for performing the methods of the invention as well as related methods for controlling the spread of viral infections and/or secondary bacterial infections.

Description

. .

Methods for Classifying Subjects Exposed to Viral Infection

[0001] Field of the Invention

[0002] The present invention relates to methods for predicting contagiousness in a human, or non-human mammal, with an RNA respiratory virus, especially influenza. In particular, the present invention provides methods for assessing whether a person is likely to become contagious with an RNA respiratory virus, such as influenza, before they show symptoms of infection. The methods of the invention may also be used for diagnosing subjects who are already contagious with such a respiratory virus, but are asymptomatic. The present invention also provides apparatus for performing the methods of the invention as well as related methods for controlling the spread of an RNA respiratory virus and/or secondary bacterial infections.

[0003] Background to the Invention

[0004] Acute upper and lower respiratory infections are a major public health problem and a leading cause of morbidity and mortality worldwide. Viruses are the predominant cause of respiratory tract illnesses and include RNA viruses such as respiratory syncytial virus, influenza virus, parainfluenza virus, metapneumovirus, rhinovirus and coronavirus (Hodinka, "Respiratory RNA Viruses",

Microbiol Spectr., 2016 Aug; 4(4)).

[0005] Influenza virus, in particular, is associated with substantial mortality and morbidity worldwide through seasonal epidemics and the occasional emergence of novel strains that lead to pandemics (Nicholson et al., "Influenza", Lancet. 2003; 362: 1733-45). The CDC estimates that in the 2015- 2016 period in the United States there were 25 million influenza illnesses, 1 1 million influenza- associated medical visits, 310,000 influenza-related hospitalizations, and 12,000 pneumonia and influenza deaths (Rolfes et al., "Estimated Influenza Illnesses, Medical Visits, Hospitalizations, and Deaths Averted by Vaccination in the United States", https://www.cdc.gov/flu/about/disease/2015- 16.htm, 2016 Dec 9 [Cited: 2017 Jan 16]). In 2003 the annual economic burden of influenza in the US alone was estimated to be around 87 billion dollars (Molinari et al., "The annual impact of seasonal influenza in the US", Measuring disease burden and costs, 28 June 2007; 25(27): 5086- 5096).

[0006] Natural influenza transmission in humans occurs over short distances, primarily via droplets.

During influenza infection, the virus is shed in nasal and pharyngeal secretions and dispersed through sneezing and coughing. A review of experimental influenza studies in volunteers found that viral shedding peaked on the 2^nd day of inoculation and stopped completely by the 6th or 7th day. Viral shedding level and duration can be reduced by treatment, but patients with seasonal influenza may be able to infect others 1 day before symptoms appear and up to 5 days after they appear (Mohamed et al., "Communicability of H1N1 and seasonal influenza among household contacts of cases in large families", Influenza and Other Respiratory Viruses, 2012; 6(3): e25-e29). [0007] Some individuals, known as "silent spreaders", may be contagious in the sense that they are able to transmit live virus to others without showing any symptoms. It is thought that influenza viral particles may be transmitted in droplets or in the form of small particle aerosols in a subject's breath. It is still unknown whether or to what degree asymptomatic individuals could transmit infection to others, although mathematical models typically assume that 33% to 50% of infections are asymptomatic or subclinical, and these individuals are around half as infectious as symptomatic cases (Lau et al., "Viral shedding and clinical illness in naturally acquired influenza virus infections", J Infect Dis., 2010; 201(10): 1509-1516).

[0008] Whilst the 2007 Guideline for Isolation Precautions: Preventing Transmission of Infectious Agents in Healthcare Settings issued by Siegel et al. and the Healthcare Infection Control Practices

Advisory Committee ("CDC 2007") indicate different modes of transmission for other respiratory viruses, transmission by droplets may be a major contributing factor to contagiousness, at least amongst adults. For instance, CDC 2007 indicates indirect contact as a primary mode of transmission for respiratory syntactical virus (RSV) amongst paediatrics for whom shared toys, for example, may become a vehicle for transmitting respiratory viruses, but the dispersal of droplets by coughing and sneezing is likely to be at least as important for adult patients.

[0009] The costs of influenza are clearly substantial and any method that could reduce the transmission of the disease would be of great benefit. If it could be determined who had increased potential of transmitting influenza virus, they could be isolated or treated, or given a mask, or any other means employed to prevent further spread of infection including preventing spread of the virus and/or secondary infections. In many situations, it may be desirable to identify individuals who are contagious or likely to become contagious for a respiratory viral infection as early as possible, before symptoms of infection are shown, in order to assist in controlling the spread of the virus. For instance, in some situations it may be valuable to screen a group of individuals who are about to spend time in close proximity to one another, for example in an aircraft cabin, to identify those members of the group who are contagious or likely to become contagious for a viral infection before they show symptoms. Individuals who are contagious or predicted to become contagious in the near future may then be treated or excluded from the group, or given a mask, or any other means employed to prevent further spread of infection. [0010] RT-PCR is the gold standard laboratory method for confirming viral respiratory infections among symptomatic individuals. However, while PCR-based diagnostics are useful in classifying infectious pathogens, they lack sensitivity. In a field study involving 322 households with index patients having influenza-like illness with symptom onset in the previous 48 hours, and who were positive for influenza A or B virus, Lau et al. estimated the peak level of RT-PCR test sensitivity to be 79.6%> [76.5%o, 83.0%o] for symptomatic infections. (Lau et al. "Inferring influenza dynamics and control in households", PNAS, 2015; 112(29): 9049-9099). The relative sensitivity of asymptomatic infections was not identifiable in this analysis and was assumed to be half that of symptomatic infections (Lau et 2010, ibid.)

[0011] Further, the proportion of infections that are asymptomatic or subclinical, and the degree to which these are contagious, as well as the proportion of shedding which occurs prior to onset of symptoms, affect the potential impact of control measures. (Lau et al., 2010, ibid.)

[0012] It has also been reported that a viral pathogen in upper airways can increase airborne dispersal of coagulase -negative staphylococci (CoNS) in nasal S. aureus carriers (Bischoff, et al., "Airborne Dispersal As a Novel Transmission Route of Coagulase-Negative Staphylococci: Interaction between Coagulase-Negative Staphylococci and Rhinovirus Infection", Infection Control and Hospital

Epidemiology, June 2004; 25(6): 504-511.

[0013] While no specific method currently exists for measuring influenza contagiousness, viral titres and clinical symptoms could provide some insight. However, measuring viral load through qPCR does not necessarily indicate how contagious a subject is and can be prone to false negative or false positive results. Symptoms such as cough or sneezing are generally reported by the subject and may not be reliable or objective. A panel of genes would be a more accurate measurement with less variation in interpretation.

[0014] Woods et al., "A Host Transcriptional Signature for Presymptomatic Detection of Infection in Humans Exposed to Influenza H1N1 or H3N2", PLOS ONE, January 2013; 8(1): e52198 describe the generation of a viral gene signature (or factor) for symptomatic influenza that is capable of detecting

94% of infected cases. The gene signature is detectable as early as 29 hours post-exposure and is reported to achieve maximum accuracy on average 43 hours (p = 0.003, H1N1) and 38 hours (p-value equals 0.005, H3N2) before peak clinical symptoms.

[0015] The viral gene signature developed from the work reported by Woods et al., ibid., is the subject of WO 2011/008349 A2, which discloses methods of identifying infectious disease infection prior to presentation of symptoms, assays for identifying genomic markers of infectious disease and methods for diagnosing the underlying aetiology of infectious disease. In particular, WO 2011/008349 A2 discloses methods of identifying a subject infected with a respiratory virus comprising determining gene expression levels of at least three genes of a gene signature from a peripheral blood cell sample of the subject and comparing the expression levels to standard gene expression levels. Specific gene signatures for RSV, influenza and rhinovirus are disclosed in addition to a gene signature for respiratory viral infection in general. WO 2011/008349 A2 also discloses methods for reducing the spread of a respiratory virus in a population by isolating a subject who is identified to be infected with a respiratory virus from the population. [0016] While Woods et al. and WO 2011/008349 A2 disclose methods for identifying a subject infected with a respiratory virus prior to presentation of symptoms, such methods are unsuited to predicting whether or not an individual will become contagious. The methodology described by Woods et al. and WO 2011/008349 A2 for elucidating the gene signatures comprehends applying sparse latent factor regression analysis to the expression levels of genes from peripheral blood samples taken from volunteers inoculated with live virus. The subjects' symptoms were recorded twice daily using the modified Jackson score and nasal lavage samples were obtained from each subject daily for qualitative viral culture and/or quantitative RT-PCR to assess the success and timing of infection. A modified Jackson score of > 6 was the primary indicator of successful viral infection, and subjects with this score were denoted as "symptomatic, infected". Viral titres from daily nasopharyngeal washes were used as corroborative evidence of successful infection using qualitative culture.

Subjects were classified as "asymptomatic, not infected (healthy)" if the Jackson score was less than 6 over five days of observation and viral shedding was not documented after the first 24 hours subsequent to inoculation. [0017] The gene signatures of Woods et al. and WO 2011/008349 A2 are therefore designed to identify subjects who are infected with virus and will develop significant symptoms (Jackson score > 6), but they are not set up to predict whether a subject is likely to become contagious with the virus. In the first place, the arbitrary symptom score of > 6 excludes asymptomatic subjects who are infected and may be contagious. Further, the Jackson symptom score involves symptoms that are unrelated to contagiousness, e.g. headache, malaise and myalgia. As a result, data from subjects with symptoms linked to contagiousness, e.g., runny nose, sneezing, cough, may also be excluded from the analysis if their total Jackson score is less than 6. The methodology disclosed by Woods et al. and

WO 2011/008349 A2 also fails to take into account the relevance of viral load to contagiousness and does not discriminate between symptomatic subjects with high and low viral titres. [0018] With regard to WO 2011/008349 A2 at least, it is also doubtful whether the gene signatures are capable of identifying a subject likely to become contagious with respiratory viral infection at an early, pre-symptomatic stage, because the gene expression signatures were evaluated at the time of maximal symptoms following viral inoculation for symptomatic subjects (and a matched time point for asymptomatic subjects). As mentioned in Woods et al., individual genes exhibit variable expression over time, and it is questionable whether gene expression levels at the time of maximal symptoms is indicative of expression levels at the pre-symptomatic stage when the ability to predict progression to contagiousness would be most valuable. In the case of influenza A, for example, the median time to peak symptoms was 80 hours post inoculation.

[0019] There remains a need for methods of identifying subjects who are likely to become contagious with a respiratory viral infection as early as possible, ideally before the subject shows any symptoms of infection. It would also be desirable to identify subjects who are already contagious with a respiratory viral infection, but are asymptomatic. Such methods would facilitate the control of the spread of respiratory viral infections and associated secondary bacterial infections.

[0020] Chen et al., [Predicting Viral Infection From High-Dimensional Biomarker Trajectories, J Am Stat Assoc. 2011 January 1 : 106(469): 1259-1279] utilises time-course gene expression array data to provide predictions of infected individuals in advance of the development of clinical symptoms. This analysis aims to determine whether subjects are infected, it does not predict whether subjects are contagious or likely to be contagious. The analysis described in Chen et al., does not account for weighting based on symptoms associated with increased viral transmission, for example symptoms associated with increased production of droplets/aerosols, nor does it incorporate weighting based on viral load.

[0021 ] Summary of the Invention

[0022] The present invention provides a method for diagnosing whether a subject is contagious with an RNA respiratory virus or predicting whether a subject will become contagious with an RNA respiratory virus, the method comprising measuring the expression levels of one or more genes selected from USP18, MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCL10, DDX58, GBP1P1, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML, PNPTl, TDRD7, and TORIB in a biological sample taken from the subject and assigning the subject to a class corresponding to the subject's degree or predicted degree of contagiousness by analysing the expression levels of the one or more genes using a classification algorithm.

[0023] The classification algorithm comprises a machine-learning derived algorithm derived prior to the steps of the claimed method by analysing measured expression levels of the one or more genes, measured at an early stage following inoculation with an RNA respiratory virus, in a population or sub-population of subjects who have been grouped in two or more classes according to their degree of contagiousness, as determined by their symptoms scores and viral loads. As explained further below and in the Examples, the classification algorithm is derived prior to the claimed method based on, for example a training data set, and so the classification algorithm does not need to be derived again each time a new subject or group of subjects is assessed and assigned to a class.

[0024] In accordance with another aspect of the present invention, therefore, there is provided a method of diagnosing whether a subject is contagious with an RNA respiratory virus or predicting whether a subject will become contagious with an RNA respiratory virus, the method comprising measuring the expression levels of one or more genes in a biological sample taken from the subject selected from MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCL10, DDX58, GBP1P1, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML, PNPTl, TDRD7, TORIB and USP18; wherein a pattern of increased expression of one or more of the genes in the sample indicates that the subject is contagious or likely to become contagious with an RNA respiratory virus.

[0025] As well as predicting whether a subject is likely to become contagious with an RNA respiratory virus, the method of the first aspect of the present invention may be useful for identifying subjects who are already contagious with a respiratory virus without showing any symptoms, i.e. "silent spreaders".

[0026] Literature references to and sequence listings for the above-mentioned genes are included at the end of this specification and incorporated herein by reference in their entirety. It will be understood that the references and sequence listings necessarily disclose specific alleles and are included by way of example only. The methods of the invention are not limited to such specific alleles, but may also be implemented using products of expression of different variants of the one or more genes.

[0027] As used herein, "contagiousness" is defined as a combination of specific clinical symptoms

(sneezing, coughing, runny nose) and viral load, indicating that a "contagious" subject is infected with an RNA respiratory virus and is susceptible to transmit the virus to another subject. Generally, contagiousness implies a sufficiently high viral load for transmission to another person and, in cases of increased contagiousness, symptoms such as sneezing and coughing that facilitate viral transmission via droplets, droplet nuclei or as airborne small particles.

[0028] The subject may be a human or a non-human mammal.

[0029] As shown in the Examples below, the genes that have been identified as being predictive of contagiousness in accordance with the present invention exhibit increased expression levels within about 48 hours after inoculation with a virus in subjects who go on to exhibit higher levels of contagiousness relative to those who do not become contagious, as defined above. This indicates the potential of the genes to identify subjects who are more likely to become contagious, and whether they are likely to be asymptomatic, or symptomatic and therefore having an increased likelihood of being contagious. Since symptoms of viral infection develop sooner in some subjects than in others, increased expression of the one or more genes according to the present invention may be predictive of contagiousness, before a subject shows symptoms of infection, or an early diagnostic indicator of contagiousness at about the same time as the subject begins to show symptoms of infection. In particular, the one or more genes according to the invention may be used to diagnose subjects who are contagious with the virus without showing any significant overt symptoms.

[0030] The present invention may be suitable for diagnosing or predicting contagiousness in a subject for any RNA respiratory virus, including respiratory syncytial virus, influenza virus, parainfluenza virus, metapneumovirus, rhinovirus and coronavirus. As described in the Examples below, the above- mentioned genes were identified by analysing the results of a study involving inoculation of patients with Influenza A H3N2 Perth/16/2009 virus. However, having regard to the specific genes identified, the majority of which are known interferon stimulating genes (ISGs), including all three RIG-like receptors, which are all involved in pathways mediated by pattern recognition receptors that are activated by all RNA respiratory viruses, including all influenza types, subtypes and strains, while others are directly anti-viral proteins (e.g. OAS2), increased expression of the one or more genes of the invention is expected to be diagnostic or predictive of contagiousness for any RNA respiratory virus. That is to say, infection with any RNA respiratory virus activates the same gene regulatory network involving increased expression of the one or more genes of the invention. A further diagnostic test may be used in conjunction with the present invention to determine which RNA respiratory virus(s) a subject has been infected with.

[0031 ] The expression levels of the one or more genes in the biological sample may be measured using any suitable method known in the art for quantifying the expression level of a gene, particularly a mammalian gene. In some embodiments, the expression level of the one or more genes may be measured by quantifying mRNA transcripts of the one or more genes according to the invention in the biological sample.

[0032] Preferably, a PCR-based method may be used such, for example, as RT-qPCR. Examples of qRT-PCR-based methods are disclosed by United States patent no. 7,101,663, the contents of which are incorporated herein by reference. An advantage of real-time PCR is its relative ease and convenience of use.

[0033] Alternatively, a gene expression microarray may be used of the kind disclosed in, for example, United States patent no. 6,040,138, the contents of which are incorporated herein by reference, in which a pool of labelled target cRNA molecules, which are obtained by transcribing double-stranded cDNA derived from the mRNA transcripts that are isolated from the biological sample and fragmenting the resulting cRNA transcripts, are hybridised to oligonucleotide probes having specific sequences that are immobilised at specific addresses on a solid support. After incubating the cRNA targets with the surface-bound probes, the arrays are washed and the labels on the targets may be used to quantify how much target is bound to any given feature on the array. The amount of a given surface-bound target cRNA is proportional to the expression level of the corresponding gene. [0034] Suitably, the biological sample may be a blood sample from the subject.

[0035] In some embodiments, the one or more genes may be selected from MAP2K6, ATF3, CXCLIO, TDRD7, DDX58 and GBP1P1. The one or more genes may be selected from USP18, CXCLIO, IFIT2, ATF3, GBP1P, CCL8, CCL2, or DDX58/IFIH1. [0036] Since IFIHl and DDX58 are functionally related (both being RIG-like receptors) and correlate strongly in terms of expression patterns, in some embodiments IFIHl may be substituted for DDX58. The concept of redundancy between clusters of genes is further illustrated in Figure 27 where CNTR1, DHX58 and CCL8 cluster together. Increased expression of any one of these genes is indicative of contagiousness, however, the overlapping biological role and expression patterns of these genes mean that the selection of one gene from this cluster can represent the group, e.g. CCL8 is selected as representative of the group, however the remaining members of the cluster can be substituted to achieve a very similar prediction of contagiousness.

[0037] In some embodiments, the one or more genes may include one or more genes selected from

MAP2K6, ATF3, CXCLIO, TDRD7, DDX58 (or IFIHl) and GBP1P1 and at least one additional gene selected from BCL2L14, CCL2, CCL8, HERC6, IFIT2, LAMP3, OAS2, PANK2, PML, PNPT1, TOR1B and USP18.

[0038] The one or more gene may be selected from USP18, CXCLIO, IFIT2, ATF3, GBP1P, CCL8, CCL2, or DDX58, for example the panel may comprise USP18, CXCLIO and IFIT2. [0039] In some embodiments, the expression level of a single one of the above-mentioned genes may be diagnostic or predictive of contagiousness. For instance, in some embodiments, the expression level of MAP2K6, ATF3, CXCLIO, TDRD7, DDX58 or GBP1P1 alone may be measured and compared with a reference level for expression of the gene. In some embodiments, the expression level of MAP2K6 or TDRD7 alone may be measured. The reference level may be a predetermined threshold level of expression for the gene that is indicative of contagiousness or predicted contagiousness or a baseline level of expression for the gene as described in more detail below. Advantageously, however, the sensitivity of the method of the invention may be improved by measuring the expression levels of a panel of two or more, or three or more, of the genes.

[0040] The one or more genes may be selected from USP18, CXCLIO, IFIT2, ATF3, GBP1P, CCL8, CCL2, or DDX58.

[0041] Accordingly, in some embodiments, the method may comprise measuring the expression level of a panel of genes comprising MAP2K6 or TDRD7 and one or more additional genes.

[0042] In some embodiments, the panel may comprise MAP2K6 and ATF3, MAP2K6 and CXCLIO, MAP2K6 and TDRD7, MAP2K6 and DDX58, MAP2K6 and GBP1P1, ATF3 and CXCLIO, ATF3 and TDRD7, ATF3 and DDX58, ATF3 and GBP1P1, CXCLIO and TDRD7, CXCLIO and DDX58,

CXCLIO and GBP1P1, TDRD7 and DDX58, TDRD7 and GBP1P1 or DDX58 and GBP1P1.

[0043] In some embodiments, the panel may comprise: MAP2K6, ATF3 and CXCLIO; MAP2K6, ATF3 and TDRD7; MAP2K6, ATF3 and DDX58; MAP2K6, ATF3 and GBP1P1 ; MAP2K6, CXCLIO and TDRD7; MAP2K6, CXCLIO and DDX58; MAP2K6, CXCLIO and GBP1P1 ; MAP2K6, TDRD7 and DDX58; MAP2K6, TDRD7 and GBP 1 PI ; MAP2K6, DDX58 and GBP 1 PI ; ATF3, CXCLIO and TDRD7; ATF3, CXCLIO and DDX58; ATF3, CXCLIO and GBP1P1; ATF3, TDRD7 and DDX58; ATF3, TDRD7 and GBP1P1 ; ATF3, DDX58 and GBP1P1 ; CXCLIO, TDRD7 and DDX58;

CXCLIO, TDRD7 and GBP 1 PI ; CXCLIO, DDX58 and GBP 1 PI ; or TDRD7, DDX58 and GBP1P1.

[0044] In some embodiments, the panel may comprise: MAP2K6, ATF3, CXCLIO and TDRD7;

MAP2K6, ATF3, CXCLIO and DDX58; MAP2K6, ATF3, CXCLIO and GBP 1 PI ; MAP2K6, ATF3, TDRD7 and DDX58; MAP2K6, ATF3, TDRD7 and GBP1P1 ; MAP2K6, ATF3, DDX58 and GBP 1 PI ; MAP2K6, CXCLIO, TDRD7 and DDX58; MAP2K6, CXCLIO, TDRD7 and GBP1P1 ; MAP2K6, CXCLIO, DDX58 and GBP1P1 ; MAP2K6, TDRD7, DDX58 and GBP1P1 ; ATF3, CXCLIO, TDRD7 and DDX58; ATF3, CXCLIO, TDRD7 and GBP1P1 ; ATF3, CXCLIO, DDX58 and GBP 1 PI ; ATF3, TDRD7, DDX58 and GBP 1 PI ; or CXCLIO, TDRD7, DDX58 and GBP1P1.

[0045] In some embodiments, the panel may comprise MAP2K6, ATF3, CXCLIO, TDRD7, DDX58 and GBP1P1.

[0046] Suitably, the method may comprise measuring the expression level of a panel of genes

comprising MAP2K6 and at least one additional gene. In some embodiments, the method may comprise measuring the expression level of MAP2K6 and a least one of ATF3, CXCLIO, TDRD7, DDX58 and GBPlP.

[0047] In some embodiments, the method may comprise measuring the expression level of a panel of genes comprising MAP2K6; ATF3 and CXCLIO, ATF3 and TDRD7, ATF3 and DDX58, ATF3 and GBP1P1, CXCLIO and TDRD7, CXCLIO and DDX58, CXCLIO and GBP1P1, TDRD7 and DDX58, TDRD7 and GBP1P1 or DDX58 and GBP1P1 ; and optionally one or more additional genes.

[0048] Alternatively, the method may comprise measuring the expression level of a panel of genes

comprising TDRD7 and at least one additional gene. In some embodiments, the method may comprise measuring the expression level of TDRD7 and a least one of MAP2K6, ATF3, CXCLIO, DDX58 and GBPlP.

[0049] In some embodiments, the method may comprise measuring the expression level of a panel of genes comprising TDRD7; ATF3 and CXCLIO, ATF3 and DDX58, ATF3 and GBP1P1, ATF3 and MAP2K6, CXCLIO and DDX58, CXCLIO and GBP1P1, CXCLIO and MAP2K6, DDX58 and GBP1P1, DDX58 and MAP2K6 or GBP1P1 and MAP2K6; and optionally one or more additional genes.

[0050] In some embodiments, the method may comprise measuring the expression level of a panel of genes comprising MAP2K6 and TDRD7 and optionally one or more additional genes. [0051 ] In some embodiments, the method may comprise measuring the expression level of a panel of genes comprising: MAP2K6 and TDRD7; ATF3 and CXCL10, ATF3 and DDX58, ATF3 and GBP1P1, CXCL10 and DDX58, CXCL10 and GBP1P1 or DDX58 and GBP1P1; and optionally one or more additional genes. [0052] In some embodiments, the method may comprise measuring the expression level of a panel of genes comprising one or more, or all of: USP18, CXCL10, IFIT2, ATF3, GBP1P, CCL8, CCL2, or DDX58.

[0053] Where one or more additional genes are included in the panel, some or all of them may be

selected from the above lists. Unless indicated otherwise, the present invention does not exclude the possibility of including within the panel one or more further genes not specifically disclosed herein, which may be found to improve further the accuracy, sensitivity or specificity of the methods of the invention.

[0054] In accordance with the present invention, a pattern of elevated expression of the one or more genes, relative to corresponding early expression levels in subjects who do not become contagious with the virus, as defined herein, is indicative of contagiousness or predicted contagiousness.

[0055] In some embodiments, the expression level of each of the one or more genes may be compared with a respective reference level. As mentioned above, the reference level may be a threshold expression level that indicates contagiousness or predicted contagiousness. Alternatively, the reference level may be a baseline level of expression which indicates that the subject is unlikely to become contagious for the respiratory virus. Significantly increased expression of the one or more genes relative to their respective baseline levels, for instance by at least l.lx, preferably at least 1.5x or 2x, may be indicative of contagiousness or predicted contagiousness.

[0056] In some embodiments, the method may involve an individual reference level for each gene.

Increased expression of at least one of the genes, preferably two or more of the genes, relative to their respective reference levels may indicate contagiousness or predicted contagiousness in accordance with the present invention.

[0057] In some embodiments, the reference level for the, or each, gene may be a previously measured expression level for the gene in the same subject. In particular, the reference level for the, or each, gene may comprise a baseline expression level of the gene for the subject which is measured at a time when the subject is known not to be infected with an RNA respiratory virus such, for example, as influenza. Where previous expression levels for the one or more genes, measured on more than one previous occasion, are available for a subject, the reference level for each gene may comprise an average of multiple previous levels. [0058] Thus, in some circumstances, a subject may be tested once to obtain baseline levels for the one or more genes, which form reference levels that may be used subsequently in case of suspected viral infection or a routine check, for comparison with contemporaneous expression levels to predict whether or not the subject is likely to become contagious with an RNA respiratory virus or to diagnose if the subject is already contagious with the respiratory virus, particularly in cases where the subject is asymptomatic. Where the contemporaneous expression levels of the one or more genes are significantly increased relative to the respective baseline levels, the subject may be contagious or likely to become contagious for the respiratory virus in accordance with the invention. Thus, in some embodiments, increased expression of the one or more genes by 1.1, 1.5, 2, 2.5, 3 or more times relative to their respective baseline levels may indicate that the subject is likely to become contagious.

[0059] In other situations, the expression levels of the one or more genes may be measured repeatedly over a period of a few days or weeks to monitor for changes in the expression levels. For instance, in some embodiments, the expression levels of the one or more genes may be monitored for at least 2 days and preferably longer, e.g., 3-10 days. Suitably, the expression levels may be measured at intervals of 1-7 days, preferably 1-3 days. In some embodiments, the expression levels may be measured every day or every other day. In some embodiments, the expression levels may be measured three times a day, twice a day, or once a day. This may be useful when the subject being tested is about to carry out or undergo an activity where it would be preferable for the subject not to be contagious with a respiratory virus. For example, the subject may be about to travel on an aeroplane in close proximity with other individuals, or the subject may be about to travel away from an area where infection with a respiratory virus is prevalent. The expression levels of the one or more genes for the subject may be measured regularly over a period of a few days prior to travel to look for the changes in the expression levels of the one or more genes that may be diagnostic or predictive of contagiousness for the virus over the ensuing few days, so that travel while the subject is contagious can be avoided. As described above, a significant increase in expression of one or more of the genes, for example by 1.1, 1.5, 2, 2.5, 3 or more times, during the period of testing may be indicative that the subject is contagious or likely to become contagious with the respiratory virus.

[0060] In other embodiments, the pattern of increased expression of the one or more genes may be

discerned using a classification algorithm. In particular, the expression levels of the one or more genes in the subject may be analysed using a classification algorithm to determine whether the subject is contagious or predicted to become contagious with the respiratory virus or not and, in some embodiments, how contagious the subject is predicted to become. Advantageously, the expression levels of two or three or more of the genes may be analysed using the classification algorithm to detect a pattern of increased expression of the genes that indicates contagiousness or predicted contagiousness.

[0061] Classification algorithms of this kind are well known in the art may be derived by machine- learning techniques using a training dataset. Expression levels for the one or more genes, measured at an early stage following infection with a respiratory virus, in a population or sub-population of subjects who have previously been grouped into two or more classes, for example using a clustering algorithm, based on their symptom scores and viral loads, may be used as a training dataset to build a classification algorithm for putting a subject into one of the classes by analysing their measured expression levels for the one or more genes. [0062] Suitably the symptom scores may comprise scores for sneezing, runny nose, and coughing, which are likely to make a subject more contagious by facilitating transmittal of live viral particles.

Examples of symptoms associated with contagiousness are: runny nose, sneezing, cough, and other symptoms that facilitate viral transmission via droplets, droplet nuclei or as airborne small particles.

[0063] The expression levels for the one or more genes may be measured up to 120 hours after

inoculation or possible exposure to RNA respiratory virus, or in some cases up to 60 hours after inoculation or possible exposure to RNA respiratory virus, for example between 24-60 or 36-60 hours after inoculation or possible exposure to RNA respiratory virus, typically around 44-52 hours, or, in particular, around 40-48 hours e.g., about 48 hours or 43 hours after inoculation or possible exposure to RNA respiratory virus. [0064] In some embodiments, the classification algorithm may comprehend two classes of subjects, namely those who are contagious or predicted to become contagious and those who are not predicted to become contagious. In other embodiments, the classification algorithm may comprise three classes: subjects who are not predicted to become contagious, subjects who are contagious or predicted to become contagious without exhibiting significant symptoms of infection ("silent spreaders") and subjects who are predicted to become more contagious by virtue of symptoms such as sneezing and coughing in addition to increased viral loads.

[0065] Numerous clustering algorithms are available to those skilled in the art for clustering subjects into two or more classes based on their symptoms scores and viral load data. Similarly, numerous machine learning techniques are available for using a training dataset comprising the two or more classes and their respective expression levels for the one or more genes to derive a classification algorithm that is able to classify a new subject by analysing their early expression levels of the one or more genes. The performance of a classification algorithm built using a machine learning process may be validated using one or more known validation methods, e.g. cross-validation, and calculating statistical parameters (e.g. accuracy, sensitivity, specificity) so that the person skilled in the art can obtain a classification algorithm that is best suited for classifying subjects by analysing their expression levels of the one or more genes.

[0066] Typically, clustering algorithms, machine learning processes and the resulting classification algorithms may be carried out using a computer.

[0067] As explained above the invention provides a method for predicting whether a subject will

become contagious with an RNA respiratory virus or diagnosing whether the subject is already contagious with the virus, the method comprising measuring the expression levels of one or more genes selected from MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCL10, DDX58, GBP1P1, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML, PNPT1, TDRD7, TOR1B and USP18 in a biological sample obtained from the subject and assigning the subject to a class corresponding to the subject's degree of contagiousness or predicted degree of contagiousness by analysing the expression levels of the one or more genes using a classification algorithm.

[0068] As described above, the classification algorithm may comprise a machine-learning derived

algorithm by analysing measured expression levels for the one or more genes, measured at an early stage, as described above, following inoculation with an RNA respiratory virus such, for example, as influenza, in a population or sub-population of subjects who have been classified according to their symptom scores and viral loads. In some embodiments, peak symptom scores and/or peak viral loads may be used. The measured expression levels for the one or more genes and the classes of subjects may be used as a training dataset to obtain the machine-learning derived algorithm. If a single decision tree provides an accurate performance, then respective threshold levels may be established for the one or more genes. Accordingly, in some embodiments, the classification algorithm may comprise respective threshold levels of the kind described above for the one or more genes.

However, if a more complex algorithm, e.g., a random forest comprising a multitude of decision trees, provides a better performance, then the whole training dataset will be used to diagnose or predict the contagiousness status of a new subject based on a pattern of elevated expression of the one or more genes.

[0069] The subjects may have been classified automatically using a clustering algorithm based on the subjects' symptoms score and viral loads following inoculation with the RNA respiratory virus. The subjects may be classified into two or more classes according to the quality of the data. Thus, in some embodiments, the subject may be classified as those who are contagious or predicted to become contagious and those who are not predicted to become contagious. In other embodiments, there may be three classes, including those who are or who are likely to become contagious but substantially asymptomatic and those who are likely to become more contagious with overt symptoms, as well as those who are not predicted to become contagious. [0070] In some embodiments, the training dataset may comprise data obtained from a sub-population of subjects, characterised by their gender, age, ethnicity, etc.

[0071] In some embodiments, the accuracy of the methods according to the present invention may be improved by combining two or more different ways of analysing the expression levels of the one or more genes. Thus, the expression levels of one or more genes, preferably two or more genes, for a subject may be compared with respective reference levels such, for example, as threshold levels or baseline levels as described above, for instance by comparing contemporaneous expression levels with one or more sets of previously obtained expression levels for the subject, and also operated on by a classification algorithm generated by machine learning techniques as described above. This may be particularly useful in situations where contemporaneous expression levels of the one or more genes are compared with one or more sets of previously obtained expression levels for the same subject, where the infection status of the subject at the time of measuring the one or more previous expression levels is unknown. The classification algorithm may therefore be used to validate the results of comparing the contemporaneous expression levels with the previously obtained levels. [0072] The present invention therefore provides a method for diagnosing whether an individual is

contagious with a RNA respiratory virus without exhibiting symptoms or predicting whether an individual is likely to become contagious with a RNA respiratory virus before the individual exhibits overt symptoms of infection or where the individual becomes contagious without exhibiting symptoms. This may be useful in controlling the spread of the RNA respiratory virus as well as associated secondary bacterial infections.

[0073] Thus, in some embodiments, an individual who has been diagnosed, or predicted, as being

contagious with a RNA respiratory virus or who is predicted to become contagious with a RNA respiratory virus may be treated to reduce their viral load or, in the case of a subject who is predicted to become contagious with symptoms, treated to ameliorate their symptoms when they materialise, particularly those that facilitate the spread of live virus particles such, for example, as sneezing and coughing.

[0074] In a third aspect of the present invention, therefore, there is provided a method for controlling the spread of infection with an RNA respiratory virus such, for example, as influenza, which comprises measuring the expression levels of one or more genes selected from MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCL10, DDX58, GBP1P1, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML,

PNPT1, TDRD7, TOR1B and USP18 in biological samples taken from each member of a group of subjects, identifying those subjects who have a pattern of increased expression of the one or more genes, indicating that they are contagious or likely to become contagious with the virus, and flagging the identified subjects for treatment or separation from the remainder of the group. Separation from the remainder of the group may include being given a mask, or any other means employed to limit the spread of the RNA respiratory virus and/or secondary infection.

[0075] As described above, a reference level of expression of the, or each, one of the genes for each individual member of the group may be used, for instance a threshold or baseline level.

Alternatively, a more complex pattern of increased expression of the one or more genes may be used to classify the subjects into more classes corresponding to their degree of contagiousness or predicted degrees of contagiousness.

[0076] Accordingly, in some embodiments, the method of the third aspect of the invention may comprise comparing the expression level of the, or each, gene for each member of the group with a previously determined baseline or threshold expression level of the gene for the same member.

[0077] The previously determined baseline expression level may be obtained by measuring the

expression level of the gene in a previously obtained biological sample from the same member. An increase in the expression level of the gene for the member relative to the baseline level of expression of the gene for the same member may indicate that the member is contagious or likely to become contagious with the viral infection. Advantageously, if the expression level of the gene for a member of the group does not show any significant increase relative to the baseline level, the newly measured expression level may be used to improve the robustness of the baseline level for future use.

[0078] Alternatively, the expression levels of the one or more genes of each member of the group may be compared with respective threshold expression levels derived by analysing the expression levels of the one or more genes at an early stage following inoculation with an RNA respiratory virus, as described above, in a population or sub-population of subjects who have been grouped in two or more classes according to their degree of contagiousness, as determined by their symptoms scores and viral loads. As described above, machine learning techniques may be used to build classification models using the expression levels of the one or more genes for the two or more classes. [0079] In a further alternative, the expression levels of the one or more genes for each member of the group may be used to classify each member in one of two or more classes according to their degree of contagiousness or predicted contagiousness using a classification algorithm derived by analysing the expression levels of the one or more genes at an early stage following inoculation with an RNA respiratory virus, as described above, in a population or sub-population of subjects who have been grouped in the two or more classes according to their degree of contagiousness, as determined by their symptoms scores and viral loads. As described above, machine learning techniques may be used to build classification models using the expression levels of the one or more genes for the two or more classes and the population or sub-population of subjects may be grouped in the two or more classes using a suitable clustering algorithm. [0080] The methods of the present invention may be carried out entirely in situ. Alternatively, expression level data may be obtained from one or more subjects and transmitted to a remote server for analysis.

[0081 ] Accordingly, in a fourth aspect of the present invention there is provided a method for controlling the spread of an RNA respiratory virus comprising measuring the expression levels of one or more genes selected from MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCL10, DDX58, GBP1P1, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML, PNPT1, TDRD7, TOR1B and USP18 in a biological sample taken from each member of a group of subjects; transmitting data encoding the expression levels for each subject to a server for analysis to look for a pattern of increased expression of the one or more genes indicative of a likelihood of becoming contagious with the RNA respiratory virus; receiving data from the remote server identifying the members of the group who are contagious or likely to become contagious and treating the flagged subjects or separating them from the remainder of the group.

[0082] In each case, increased expression of the, or each gene, relative to a respective reference level may indicate likelihood of becoming contagious for the RNA respiratory virus. Alternatively, a pattern of expression of two or three or more of the genes may be used by a classification algorithm of the kind described above to assign the subject to one of two or more classes that are characterised by different degrees of contagiousness or predicted degrees of contagiousness, including not predicted to become contagious. [0083] The methods of the present invention may be carried out using apparatus and equipment of the kind that is known and available to those skilled in the art, including PCR equipment, gene expression microarrays, suitable blood sampling equipment and data processing apparatus (i.e.

computers). It is envisaged that the methods of the invention may be carried out entirely in situ, but in some embodiments, different parts of the method may be carried out in different locations, with data being transmitted between those different locations via the Internet or a wide area network.

Thus, for example, blood samples may be taken from one or more subjects in a first location. The blood samples may be analysed to measure the expression levels of the one or more genes at the first location, or the blood samples may be transferred physically to a second location for analysis. The expression levels of the one or more genes may be analysed to determine whether any of the one or more subjects are contagious or predicted to become contagious for a RNA respiratory virus at the same location where measurement of the expression levels is performed, or data encoding the expression levels may be transferred to a third location for analysis. [0084] Accordingly, in a fifth aspect of the present invention, there is provided networked apparatus for determining whether a subject is at risk of becoming contagious with an RNA respiratory virus such, for example, as influenza, the apparatus comprising:

(a) gene expression measuring equipment that is operable to measure the expression levels of one or more genes in a biological sample taken from the subject selected from MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCL10, DDX58, GBP1P1, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML, PNPT1, TDRD7, TOR1B and USP18 and to encode the measured expression levels as expression level data associated with identity data identifying the subject;

(b) a server that is operable to receive the expression level data and identity data and to execute program code that analyses the expression levels of the one or more genes for a pattern of increased expression that is indicative that the subject is contagious or predicted to become contagious for the RNA respiratory virus, and generates contagiousness data associated with the subject, the contagiousness data indicating whether or not the subject is contagious or likely to become contagious with the RNA respiratory virus; and

(c) one or more electronic communication components for connecting the measuring equipment to the server to allow the transmission of the expression level and identity data from the measuring device to the analyser.

[0085] The server is suitably operable to execute the program code to analyse the expression levels of the one or more genes in accordance with the preceding aspects of the present invention.

[0086] Subjects who are diagnosed as being contagious with an RNA respiratory virus or predicted to become contagious with such a virus in accordance with the methods or using the apparatus of the present invention may be quarantined to prevent the spread of the virus. Alternatively, such subjects may be treated to alleviate their viral load and/or symptoms of the virus that are susceptible of spreading the virus such, for example, as sneezing, runny nose or coughing.

[0087] In some embodiments, therefore, subjects who are diagnosed as being contagious or predicted to become contagious may be treated with an anti-viral agent. Numerous anti-viral agents are available to those skilled in the art. In the case of influenza, the anti-viral agent may, by way of example, be selected from amantadine, rimantadine, oseltamivir and zanamivir. For RSV, ribavirin may be used for example.

[0088] Alternatively, or in addition, subjects who are diagnosed as being contagious or predicted to become contagious, particularly those who are predicted to become contagious with symptoms, may be treated with one or more medicaments for alleviating symptoms of infection such, for example, as sneezing or coughing. Numerous medicaments are available to those skilled in the art for treating sneezing and runny nose, including saline solutions administered in the form of nose sprays or mist, topical nasal decongestants and oral nasal decongestants (e.g. phenylephrine and pseudoephedrine). Similarly, numerous medicaments are known to those skilled in the art for alleviating coughing, including oral cough suppressants (e.g. codeine, hydrocodone, dextromethorphan and

diphenhydramine), oral expectorants (e.g. Guaifenesin), and topical medicines such as camphor and menthol. Cough treatments may be administered for example in the form of lozenges or drops.

[0089] Given that persons who have symptoms of a respiratory viral infection such, for example, as sneezing, coughing or runny nose may also facilitate the transmission of secondary bacterial infections, in some embodiments, subjects who are identified as being contagious or predicted to become contagious with a respiratory viral infection, especially those who are predicted to become contagious with symptoms of sneezing, runny nose and/or coughing, may be administered a suitable anti-biotic treatment. Numerous antibiotics are available to those skilled in the art, including penicillins, tetracyclines, cephalosporins, quinolones, lincomycins, macrolides, sulphonamides, glycopeptides, aminoglycosides and carbapenems. By way of example only, a subject who is predicted to become contagious may be prescribed course of treatment with amoxicillin, doxycycline, cephalexin, ciprofloxacin, clindamycin, metronidazole, azithromycin,

sulfamethoxazole/trimethoprim, amoxicillin/clavulanate or levofloxacin. Such treatment may be especially appropriate for subjects who are already known to have a bacterial infection or carrying a pathogenic bacterium. [0090] The invention also provides a classification algorithm for assigning a subject to a class

corresponding to the subject's degree, or predicted degree, of contagiousness, wherein the classification algorithm is based on measured expression levels of one or more genes, measured at an early stage following inoculation with an RNA respiratory virus, in a population or sub-population of subjects who have been grouped in two or more classes according to their degree of contagiousness, as determined by their symptoms scores and viral loads.

[0091] The classification algorithm may analyse the expression levels of one or more genes selected from USP18, MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCL10, DDX58, GBP1P1, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML, PNPT1, TDRD7, and TOR1B.

[0092] The classification algorithm may be computer- implemented and comprise receiving in a

computer a data set comprising expression levels of one or more genes from one or more subjects and executing on the computer software to classify the one or more subjects according to their degree, or predicted degree, of contagiousness.

[0093] The classification algorithm comprises a Naive Bayes classifier, a support vector machine

classifier or a decision tree, or preferably a random forest classifier. [0094] The invention includes a computer-readable medium and/or computer program comprising instructions which, when executed by a computer, cause the computer to carry out the classification algorithm of the invention.

[0095] Brief description of the drawings [0096] Following is a description by way of example only with reference to the accompanying drawings of embodiments of the present invention.

[0097] In the drawings:

[0098] FIG. 1 is a chart showing volunteer visits for a clinical study as described in Example 1 below.

[0099] FIG. 2 is a scatterplot of VAS and categorical data. The dots represent the intersection of the VAS and categorical data for the same patient at the same time point. The lines show the intersection where VAS and categorical data are comparable. If both scales are comparable, it is expected that dots fall into the lines. This is not the case since the dots often fall below the lines for 2 and 3 in the categorical scale.

[00100] FIG. 3 shows VAS scores for cough for days -1 to 8 for individual subjects. [00101] FIG. 4 shows VAS scores for sneezing for days -1 to 8 for individual subjects.

[00102] FIG. 5 shows VAS scores for runny nose for days -1 to 8 for individual subjects.

[00103] FIG. 6 shows qPCR viral shedding values for days -1 to 8 for individual subjects.

[00104] FIG. 7 is a correlation matrix of variables for symptoms and qPCR viral shedding data. Darker intersections indicate highly correlated values whereas lighter intersections indicate lower correlation. [00105] FIG. 8 is a scree plot for k-means algorithm.

[00106] FIG. 9 show clustering of subjects based on k-means method. Note that the clustering was

performed with scaled values but original values are shown in the figure.

[00107] FIG. 10 is a density plot of gene expression microarrays after RMA background correction and quantiles normalization. [00108] FIG. 11 is a heatmap of the average values for three levels of contagiousness based on the results of the microarray.

[00109] FIG. 12 is an importance plot of the variables included in the random forest classification model based on the results of the microarray. [00110] FIG. 13 comprises violin plots showing the accuracy of the models with 2 to 10 most important variables and with the 35 variables based on the results of the microarray.

[00111] FIGS. 14A-14T show expression data of selected genes from day -1 to day 8 from the microarray.

Each of the points represents a time point of a subject, with level 3 being the subjects with high virology and high symptoms (symptomatic), level 2 having high virology but lower symptoms

(asymptomatic) and level 1 having low values for both. The lines show the average values of the samples per group. The dotted line indicates the time point before which the analysis was performed. The graphs demonstrate the separation in subject groups at day 2 (am) for these genes, indicating a potential to predict which subjects are more likely to become contagious and whether they are asymptomatic or symptomatic and therefore having increased likelihood of being contagious.

[00112] FIG. 15 illustrates schematically one of the mammalian signalling pathways involving pattern recognition receptors that are activated by RNA respiratory viruses (Katze et al., "Innate immune modulation by RNA viruses: emerging insights from functional genomics", Nature Reviews

Immunology, 2008; 8: 644-654). [00113] FIG. 16 is a flowchart of a method in accordance with one embodiment of the present invention.

[00114] FIG. 17 is a schematic illustration of networked apparatus according to the present invention

[00115] FIG. 18 is a flowchart of a method according to another embodiment of the invention, which may be carried out using the networked apparatus of FIG. 17.

[00116] FIG. 19 is a flowchart showing the derivation and use of a classification algorithm for classifying a subject according to their predicted degree of contagiousness with an RNA respiratory virus, e.g. influenza.

[00117] FIGs. 20 to 25 are plots to show qPCR results for the top 6 genes identified in Table 2Time in days is shown on the x-axis and the negative Delta CT value on the y-axis. The graphs demonstrate the differential expression between the three contagiousness groups - low, medium and high. [00118] FIG. 26 shows the time course of the study in Example 1. Time 0 is the time of inoculation.

Times at which the PAXGene™ RNA samples (used for transcriptomics), nasal swabs (qPCR to measure virology) and the symptom diary cards were utilised are demonstrated.

[00119] FIG. 27 shows principle components analysis based on the microarray results.

[00120] Examples [00121] Example 1 [00122] As described below, subjects were clustered into three levels of contagiousness based on their virology levels and clinical symptoms. Affymetrix™ HG-U133 Plus 2.0 microarray chips for time points from day -1 to day 2 in the morning were used to perform transcriptomics analysis. Time course analysis revealed 19 genes that are more expressed in early time points after inoculation in subjects with higher levels of contagiousness. This could be used to distinguish between noncontagious, low contagiousness and highly contagious. "Contagiousness" was defined as a combination of specific clinical symptom and viral shedding data.

[001231 Methods

[00124] 60 healthy volunteers were inoculated intranasally with influenza A H3N2 Perth/16/2009. All volunteers provided informed consent and underwent extensive pre-enrolment health screening

(FIG. 1) and any with significant baseline antibodies to the strain of influenza utilised were excluded.

After approximately 48 hours in quarantine (approximately mid-day on study day 0), a predetermined dose of influenza A was instilled into bilateral nares of subjects using standard pipetting methods.

The volunteers had clinical measurements and samples taken until discharged from quarantine and then at each follow up visit.

[00125] 34 of the 60 subjects became infected after inoculation (evidenced by confirmed viral shedding), 24 were identified as not infected and 2 inconclusive. An interim analysis was performed after the first 27 were inoculated and all samples for each subject were sent for gene microarray assays. One of the 27 subjects did not complete the quarantine and so was excluded from analysis. Of the 26 subjects with viable microarray data, 12 were identified as confirmed as infected and 14 as not infected.

[00126] At predetermined intervals, blood was collected into RNA PAXGene™ collection tubes (once on Day -1, then twice daily thereafter on days 0 to 6 and once daily on days 7, 8, 15 and 28). Epithelial lining fluid was collected from nasopharyngeal FLOQ swabs twice daily (starting on Day 1 morning, first sample approximately 20 hrs post inoculation). Blood and nasal collection continued throughout the duration of the quarantine. Sample collection and timings are represented in Figure 26.

[00127] Subjects self-assessed their symptoms three times daily throughout quarantine on both categorical and continuous (Visual Analogue Scale, VAS) symptom diary cards. Categorical symptoms were recorded using a modified standardized symptom score. The modified Jackson Score requires subjects to rank 12 symptoms consisting of: upper respiratory tract symptoms (runny nose, stuffy nose, sore throat, sneezing, and earache), lower respiratory symptoms (cough, shortness of breath, and wheeze) and systemic symptoms (headache, myalgia, muscle and/or joint aches,

chilliness/feverishness) on a scale of 0-3 of "no symptoms", "just noticeable", "bothersome but can still do activities" and "bothersome and cannot do daily activities". Additionally, shortness of breath at rest and wheeze at rest were also recorded using an additional grade for these symptoms only (grade 4 = symptoms at rest). VAS symptoms were on a 0 to 10 scale.

[00128] The two scales were compared with each other, and it was determined that the VAS scale data were superior than the categorical scale data for the purposes of identifying genes in accordance with the present invention that are more expressed in early time points after inoculation in subjects with higher levels of contagiousness as defined above (see FIG. 2). Specific clinical symptoms related to contagiousness included cough (FIG. 3), sneezing (FIG. 4) and runny nose (FIG. 5).

[00129] Epithelial lining fluid was assessed by qualitative viral culture and quantitative influenza RT- PCR for success and timing of viral infection, as well as viral quantification. Viral shedding data was collected in the form of qPCR data (FIG. 6).

[00130] Pearson's correlation coefficient was calculated to explore the correlation between the individual symptoms and qPCR data (see FIG. 7). The peak value, which is the maximum value of each variable for each subject from day -1 to day 8, was selected. Sneezing and runny nose show a high correlation, while cough is not so well correlated with the former two. However, this may be due to the fact that only two subjects showed cough. The peak of the symptoms score is a composite score and was calculated as the sum of the peaks of cough, sneezing and runny nose for each subject.

Symptoms scores do not show close correlation with qPCR data.

[00131 ] Subj ects were grouped using the sum of the peak symptoms scores and the qPCR data by k- means clustering. The optimal number of clusters was determined by a scree plot (see FIG. 8) and subjects were divided into three clusters as follows:

[00132] Level 3 subjects (RVL002, RVL007, RVL024) were deemed to be the most contagious as they had high virology and symptoms.

[00133] These were followed by Level 2 subjects (RVL006, RVL009, RVL014, RVL017, RVL021, RVL022, RVL026) with low symptoms and high virology; and [00134] Level 1 subjects (RVLOOl, RVL003, RVL004, RVL005, RVL008, RVLOIO, RVL011, RVL012,

RVL013, RVL016, RVL018, RVL019, RVL020, RVL023, RVL025, RVL027) with low symptoms and low virology (FIG. 9).

[00135] Blood was assessed for gene expression utilising Affymetrix HG-U133 Plus 2.0 microarray chips, which were used to measure the transcripts' expression as described in more detail in Example 2 below. Microarray data was pre-processed using RMA background correction and quantiles normalization (see FIG. 10). [00136] Time course analysis from data from days -1 to 2 in the morning was performed in order to detect the genes showing different patterns between the defined levels of contagiousness. Statistical analysis was performed using the R package maSigPro, a tool for multi-series time -course analysis (Conesa et al., "masigPro", Bioinformatics, 2006; 22 (9): 1096-1102). The method is a two-regression step approach: the first regression fit adjusts a global model in order to select differentially expressed genes and, in the second step, a variable selection strategy is applied to identify statistically significant time profiles between the experimental groups.

[00137] A subset of genes with different profiles was identified. Most of the selected genes show a

relevant pattern (see FIG. 11 which was generated for exploratory purposes), showing separation of patients based on their cluster at day 2 (am). 33 genes were detected at the lowest threshold (0.45) with 10 genes at the mid threshold (0.5) and 3 at the highest threshold. The desired pattern has an increasing expression during the early time points at higher levels of contagiousness. GBP1P1 is an example of a gene that follows this pattern. FKBP5, DDIT4, DIP2A and TWIST2 do not show this pattern. The aim of the analysis was to determine the earliest time point at which gene expression significantly varied between different clusters to identify a list of genes that show potential early diagnosis of contagiousness or prediction of the subjects more likely to become contagious and distinguish between those who are likely to have low or high levels of symptoms. Data for each gene are shown in FIGS. 14A-14T.

[00138] Random forest was used to build a classification model able to distinguish between contagious (Levels 2 and 3) and not contagious (Level 1) subjects at day 2 in the morning based on the data of the probes selected. Models were validated using Leave-One-Out cross validation. Performance measurements for this model are 88% accuracy, 80% sensitivity, 93.33%) specificity, 88.89%) positive predictive value, 87.50%> negative predictive value and 86.67%> balanced accuracy.

[00139] The importance of the variables in this model are shown in FIG. 12. It will be observed that MAP2K6 gene shows a high importance in comparison to the other variables. However, the results of the modelling are slightly different each time the analysis is run due to strong correlation of the genes. For example PCA analysis (FIG. 27) demonstrated similarities between CCL8, CMTR1 and DHX58 yet the importance demonstrated in FIG.12 indicates CCL8 has greater predictive power. Thus the importance was only used as a measure to exclude genes with consistently low predictive power as opposed to demonstrating particular genes with the greatest predictive power. The analysis was re -run several times in order to obtain a good estimation of the accuracy of the models and to test the performance with a smaller number of genes (see FIG. 13). The first 10 variables sorted by importance are MAP2K6, TDRD7, DDX58, GBP1P1, CXCL10, ATF3, IFIH1, PML, CCL2 and PNPT1. The accuracy is similar with fewer variables. [00140] Biological analysis was also used to determine whether there was obvious redundancy in the list of genes based on functional groupings. The threshold at which the gene was detected was also taken into account as well as correlation analysis of the genes. For example, PML and SP100 are both major components of nuclear bodies; however, PML was detected at a higher threshold and shows clearer separation between severity clusters at 2am.

[00141] Based on the modelling work and the biological interpretation of the list of genes, the following genes were selected: DDX58, CCL2, CCL8, CXCL10, TOR1B, PML, OAS2, PNPT1, ATF3, GBP1P1, MAP2K6, LAMP3, TDRD7, USP18, HERC6, IFIT2, PANK2, IFIH1 and BCL2L14. Performance measurements of this model were the same as using all the probes. [00142] As noted above, it has been found that model as hereinbefore described has a sensitivity of at least about 80%. Further, surprisingly, it has been found that measuring the expression level of more than two or three of the genes may not add significantly to the predictive sensitivity of the model.

[00143] Pattern recognition receptors recognise pathogen-associated molecular patterns (PAMPs) (e.g., bacterial cell wall and viral RNA structures). Two major classes of pattern recognition receptors are the Toll-like receptors (TLRs) and RIG-like receptors (RLRs). RIG- like receptors almost specifically bind RNA structures. Crucial Toll-like receptors for influenza are TLR7, RIG-1 and potentially TLR3. All three of these receptors are capable of recognising influenza RNA and start a signalling cascade that leads to the activation of NFKB, and IRF3 and IRF7 as shown in FIG. 15.

[00144] IRF3 and IRF7 are transcription factors that cause the production of Type-I interferons (mainly alpha and beta) and Type-Ill interferons (lambda). These then bind to receptors that stimulate the

JAK/STAT pathway that leads to the expression of the interferon stimulator genes (ISGs).

[00145] The majority of the selected genes are known ISGs, and include all three of the RLRs (known positive feedback loop), as well as some directly anti-viral proteins such as OAS2.

[00146] Since the same pathways are involved in the biological response to all RNA respiratory viruses, the selected genes are expected to be diagnostic or predictive for contagiousness, not only for influenza A, but other types and subtypes of influenza and other RNA respiratory viruses such, for example, as respiratory syncytial virus, parainfluenza virus, metapneumovirus, rhinovirus and coronavirus.

[00147] Example 2 [00148] In one embodiment of the present invention, a group of individuals are planning to travel together and/or spend time in close proximity with one another. For instance, the group of individuals may comprise members of a sports team who are planning to travel by aeroplane to an event at a distant location and then to spend time together in close proximity in accommodation provided for the team. Prior to travel, there is a concern that one or more members of the team may become contagious for influenza and that if they travel and stay with the rest of the team, healthy individuals may also become infected. It is therefore desired to identify any members of the team who are contagious or likely to become contagious around the time of travel or during the event, so that they may be excluded from the team or arrangements made for them to travel separately. It is necessary to identify contagiousness before any symptoms are shown, because individuals infected with a RNA respiratory virus are often contagious before they exhibit symptoms, and, as noted above, some individuals are contagious without exhibiting symptoms. The likelihood of each member of the group becoming contagious with influenza may be predicted in accordance with the methods of the present invention.

[00149] FIG. 16 shows a flowchart for a method in accordance with the invention in which each member of the group is tested periodically over a period of several days prior to the time of travel to measure the expression levels of a panel of at least three genes which are predictive of contagiousness with an RNA respiratory virus in accordance with the present invention. The three genes in the panel may be MAP2K6, TDRD7 and DDX58. In an alternative embodiment, IFIH1 may be substituted for

DDX58. The genes in the panel may be one or more, two or more, three or more, four or more, five or more, six or more, seven or more, or all of: USP18, CXCL10, IFIT2, ATF3, GBP1P, CCL8, CCL2, or DDX58. In other embodiments, the panel may comprise some or all of the genes grouped above and two genes, or four or more genes, selected from the genes disclosed previously herein. [00150] Ideally, the first test (step 12) is made at least 10 days before the date of travel, and each member of the group is tested every other day, or every day, including the day of travel, or the day before. Generally, the first test must be made at least two days before the date of travel, with a follow-up test the day of travel or the day before.

[00151] For each test, a biological sample, typically a whole blood sample, is taken from a subject and collected in an RNA PAXGene™ collection tube. The expression levels of the genes of the panel are measured using a gene expression microarray. A suitable microarray is the HG-U133 Plus 2.0 microarray chip which is commercially available from Affymetrix Inc. (Santa Clara, USA).

[00152] Total RNA is first isolated from the biological sample using methods known to those skilled in the art. The amount of mRNA present in the isolated total RNA is then amplified to a quantity sufficient for effective hybridisation and detection. Depending on the amount of RNA isolated, one or two rounds of amplification may be required.

[00153] cDNA is generated from polyadenylated transcripts through the use of an oligo(dT) primer that binds to the mRNA poly-A tail and primes reverse transcription into a DNA/RNA hybrid molecule using reverse transcriptase. DNA polymerase I and random primers in combination with RNase H are then used to replace the RNA strand with DNA, leading to the synthesis of double-stranded cDNA. The amplification step follows, with the cDNA serving as a template for T7 RNA polymerase, leading to the production of RNA molecules from each cDNA molecule. Biotinylated nucleotide precursors are included in the reaction, so that the resulting copy RNA (cRNA) is internally labelled with biotin. If the first round of cRNA production does not lead to adequate amounts of cRNA, then the process can be repeated (cRNA to cDNA to second-round cRNA). The cRNA is then fragmented by heating to produce ~50bp segments, which is optimal for interaction with surface-bound probes and successful consistent hybridisation.

[00154] The cRNA fragments are then incubated with spike -in controls and buffers as required with a microarray comprising 25-mer cDNA probes immobilised at specific addresses on a suitable solid support. The cDNA probes comprise selected fragments of the MAP2K6, TDRD7 and DDX58 genes, or a different panel of preferred genes optimised for hybridisation to the target cRNAs. As is known by those skilled in the art, the selection of the best set of well-matched sensitive and specific probes is important for the accurate measure of transcript abundance. [00155] The arrays are incubated for 16 hours at control temperature with mixing to allow hybridisation of the labelled target cRNA fragments with the surface-bound probes on the microarray surface. After incubation, the hybridisation cocktail is removed, the array washed, and the bound biotinylated targets are stained with streptavidin (biotin binding) complexed with the fluorescent molecule phycoerythrin. The streptavidin- phycoerythrin antibody complex (SAPE) provides a fluorescence- based "report" of how much transcript is bound to any given feature on the array. The fluorescent report may be amplified by subsequent staining with biotinylated goat anti- streptavidin antibody and then again with phycoerythrin-labelled streptavidin antibody. The amount of fluorescence on the array is assessed before and/or after amplification with a laser scanner.

[00156] The scanned image of the hybridised array may be stored as a pixel-based image file, and this image is processed to determine the amount of fluorescence signal at each of the probes. These data are then processed in the manner known to those skilled in the art to derive expression levels of the genes corresponding to the probes.

[00157] As mentioned above, each subject is re -tested periodically over the period prior to the date of travel (step 14) using the same procedure as described above for the first test. After each test, the expression levels of the genes in the panel are compared with the corresponding expression levels obtained in the previous test(s) (step 16). An increase in the expression levels of one or more, preferably at least two or three, of the genes of the panel indicates that the subject may be about to become contagious for influenza (step 18). Where a subject is tested three or more times at suitable intervals, a progressive increase in the expression levels of the one or more genes may be treated as indicative that the subject will become contagious for influenza.

[00158] As known by those skilled in the art, a person infected with influenza becomes contagious about 24-72 hours after contracting the virus and remains that way for up to 5 days after the onset of symptoms. Children or people with compromised immune systems may be contagious to those around them for up to 2 weeks. Accordingly, a member of the group who is predicted to become contagious with influenza as a result of testing in accordance with the method of the present invention may be treated and/or isolated from the rest of the group for a period of at least seven days and, in some instances, 14 days, or given a mask, or other means employed to limit spread of the virus and/or secondary infection (step 20).

[00159] For instance, a member of the group who is predicted to become contagious with influenza may be treated with an anti-viral agent and/or a treatment to alleviate the symptoms of influenza. Suitable anti-viral agents and treatments are described above.

[00160] Example 3 [00161] With reference to FIG. 17, in another embodiment of the present invention, networked apparatus for testing one or more individuals 32 to predict whether or not they are contagious or likely to become contagious with an RNA respiratory virus, or are already contagious with the virus without showing symptoms, comprises a quantitative real-time-PCR machine 40 that is connected to a local computer 42 for the transmission of data therebetween. The computer 42 is connected to the Internet 44 or another wide area data communication network for communication with the remote server 48.

The server 48 comprises or is connected to a permanent memory device 50 such as a suitable hard disk drive which contains data encoding the baseline expression levels of a panel of genes for the individuals 32, as described in more detail below. The server 48 further comprises a transient memory 52 and a processor (not shown). The transient memory 52 stores executable program code in the conventional way.

[00162] In the present embodiment, the individuals 32 have previously been tested for their expression levels of the panel of genes, and the expression levels stored as baseline expression levels in the permanent memory device 50 of the server 48 with suitable identifiers of the individuals 32. After testing the individuals 32 for their baseline expression levels, the individuals were monitored to check if they developed symptoms of RNA respiratory virus infection. If they did, the data were discarded, and the individuals were re -tested once they were symptom-free. In this way, the baseline expression levels were validated as being the true expression levels of the genes when the individuals are uninfected. [00163] In circumstances where one or more of the individuals 32 wish to travel in a confined space with other individuals, for example on an aeroplane, or in a situation where one or more of the individuals 32 are located in a place where there is a risk of infection with an RNA respiratory virus, for instance in the case of a localised influenza outbreak, it may be desirable to ascertain whether the one or more individuals 32 concerned are contagious without showing any symptoms or likely to become contagious in order to control the spread of the virus and associated secondary bacterial infections.

[00164] As shown in FIG. 17, a blood sample is taken from each individual 32, as indicated within the circle marked A, and the blood samples from the one or more individuals 32 are transferred into a standard multi-well array 34 for testing. [00165] The blood samples of the one or more individuals 32 are then transferred to the qRT-PCR 40 machine for rapid measurement of the expression levels in each sample of the genes of the panel. The genes in the panel may be one or more, two or more, three or more, four or more, five or more, six or more, seven or more, or all of: USP18, CXCL10, IFIT2, ATF3, GBP1P, CCL8, CCL2, or DDX58. In the present embodiment, the genes are MAP2K6, ATF3, CXCL10 and GBP1P1, but as described in Example 2 above, different genes may be used within the panel as hereinbefore described, e.g. in other embodiments, the panel may comprise some or all of the genes grouped above and one gene, two genes, or four or more genes, selected from the genes disclosed previously herein.

[00166] The measured expression levels of the genes of the panel in each sample are encoded in suitable data files and transferred to the local computer 42, together with identity data identifying each of the one or more individuals 32, so that the expression data from each sample is associated with the correct individual. The identity data also includes a field representing the status of the individual 32 as contagious or non-contagious. At the start of the test, each individual 32 is indicated to be noncontagious.

[00167] The expression level data and identity data are then communicated from the local computer 42 to the remote server 48 via the Internet 44. The operations performed by the remote server 48 are illustrated schematically in FIG. 18.

[00168] As a first step (step 112), the server 48 receives the expression level data and identity data from the local computer 42. Using the identity data for each individual 32, the server 48 looks up the stored baseline data for the individual 32 (step 114). The server 48 then executes program code stored in its transient memory 52 to compare the baseline data for the individual 32 with the received expression level data (step 116). Using the baseline data as a reference level for each gene in the panel, the server 48 determines whether the expression level of each gene is increased relative to the reference level. [00169] If the expression level for one or more of the genes in the panel is determined to be increased relative to the baseline level, the individual 32 is flagged as being at risk of becoming contagious for an RNA respiratory virus, and the identity data associated with the individual 32 is modified accordingly to show that the individual 32 is "contagious". [00170] If the expression levels for all of the genes of the panel are determined not to be increased relative to their corresponding baseline levels, the individual 32 is not flagged as being at risk of becoming contagious. The measured expression levels for the genes in the panel may then be added to the stored baseline data for the individual to improve the quality of the baseline data for future use.

[00171] The identity data, modified if appropriate, may be transmitted or made accessible to the local computer 42, or the server 48 may initiate the transmission of an electronic communication to the local computer or another computer device (e.g., a handheld device) that is able to communicate with the server 48 via the Internet 44 to notify an operator of the results of the tests for the one or more individuals 32. If any individuals are identified as being at risk of becoming contagious, they may then be treated with a suitable anti-viral agent or medicament for alleviating the symptoms of viral infection and/or quarantined from the rest of the group. Further, in order to prevent spread of secondary bacterial infections, individuals identified at risk of becoming contagious may also be administered a course of antibiotics.

[00172] As a variation of the present embodiment, the results of the analysis may be validated using a classification algorithm of the kind described in Example 4 below, operating on the expression levels of the genes in the panel.

[00173] Example 4

[00174] Examples 2 and 3 above described embodiments of the present invention in which the measured expression levels of a panel of genes are compared with respective reference or baseline levels to determine a pattern of increased expression of the genes that is indicative of contagiousness or predicted contagiousness for an RNA respiratory virus. As disclosed herein, the measured expression levels of the panel of genes may alternatively be inputted into a classification algorithm built using machine-learning techniques for classifying a subject as to their degree of contagiousness or predicted degree of contagiousness based on the measured expression levels of the genes in the panel.

[00175] In accordance with the present invention, a classification algorithm may be derived using known machine-learning techniques based on a training dataset comprising expression levels for a population or sub-population of subjects, measured at an early stage after inoculation with an RNA respiratory virus, and data assigning the subjects into two or more classes according to their degree of contagiousness as determined by their symptom scores and viral loads. [00176] The classes may be established automatically using known clustering algorithms.

[00177] In the present embodiment, therefore, as described in Example 1 above, healthy volunteers are inoculated with an RNA respiratory virus such, for example, as influenza A H3N2 Perth/16/2009 and their symptoms scores and viral loads measured over the following days. Blood samples are taken on day 2 (am), 48 hours after inoculation, and the expression levels for a panel of three genes comprising

TDRD7, CXCLIO and GBP 1 PI measured using qRT-PCR, as described above in Example 3. The genes in the panel may be one or more, two or more, three or more, four or more, five or more, six or more, seven or more, or all of: USP18, CXCLIO, IFIT2, ATF3, GBPIP, CCL8, CCL2, or DDX58. In other embodiments, the panel may comprise some or all of the genes grouped above and one gene, two genes, or four or more genes, selected from the genes disclosed previously herein. The genes in the panel may be any of the 19 genes identified in Example 1.

[00178] Although a panel of three genes is used in this example, it will be understood that a panel

comprising one gene, two genes, three genes, or more than three genes, selected from those mentioned above may be used. In some embodiments, the panel of genes may comprise genes in addition to those listed herein.

[00179] In this embodiment, qRT-PCR is used for measuring the early expression levels of the genes in the panel, but in other embodiments different techniques may be employed for measuring the gene expression levels such, for example, as expression microarrays such, for example, as described in Example 2 above. [00180] Using k-means clustering, the subjects were grouped into two clusters based on their peak

symptoms scores and viral loads. The clusters were "contagious" and "non-contagious". Although the subjects were grouped into two clusters in this example, in other embodiments, they may be grouped into three, or even more, clusters, for example as described in Example 1 above.

[00181 ] With reference to FIG. 19, using the early expression levels of the three genes in the panel and the classification of the subjects derived using the clustering algorithm mentioned above as a training dataset 201, classification models are built using standard machine learning algorithms 202.

Advantageously, the performances of the different classification models are calculated using known validation methods 203. The best model is then selected 204 as the classification algorithm.

[00182] The classification algorithm is then used to predict whether a new subject of unknown class is likely to become contagious with influenza or diagnose whether a new subject is already contagious with influenza without showing any or significant overt symptoms.

[00183] In accordance with the present embodiment, the expression levels of the panel of three genes for the new subject are measured using qRT-PCR and inputted 205 to the classification algorithm 206. In practice, the classification algorithm may be implemented in the form of software that can be executed by computer.

[00184] The classification algorithm operates on the data to assign a class (contagious/predicted to be contagious/not contagious) to the new subject 207 and the subject is assigned to that class 208. [00185] Example 5

[00186] RT-PCR was performed to confirm the microarray results in Example 1. These experiments were based on the same RNA samples used in Example 1. Samples were re-checked for integrity and converted to cDNA before the genes of interest were measured using TaqMan Array Micro Fluidics Card (Thermo Fisher Scientific). The probes used are shown in Table 1. [00187] Data was normalised to the 18S RNA control for each sample and the negative-delta CT

calculated.

[00188] For the analysis of the results, each subject had their data normalised to the Day -1 time point.

The data from day 2 in the morning (2AM) was used to build the classification models using random forest with leave-one-out cross validation. The important variables in each iteration of the leave one out analysis were determined, highlighting five or six genes using the PCR probes with strong predictive power (Table 2). However it should be noted that due to the similar profile of all 19 genes that this should not exclude the other genes. If two genes have similar profiles, with one having slightly more predictive power than the other, one will appear as important while the other will not. FIG. 27 illustrates the concept of redundancy between clusters of genes, in this example CMTR1 , DHX58 and CCL8 cluster together. Increased expression of any one of these genes is indicative of contagiousness, however, the overlapping biological role and expression patterns of these genes mean that the selection of one gene from this cluster can represent the group, e.g. CCL8 is selected as representative of the group, however the remaining members of the cluster can be substituted to achieve a very similar prediction of contagiousness. The profiles of the six genes highlighted by the analysis are demonstrated in Figures 20 to 25. As noted above, any of these six genes can be substituted and/or combined with any of the 19 genes identified in Example 1. In particular, MAP2K6 and TDRD7 show high "importance" values based on transcriptomics alone (FIG. 12), and so these genes can be combined or substituted with any of the six genes identified as having the strongest predictive power based on qPCR results. [00189] The validation metrics comparing an analysis in which all the genes are measured with the

analysis in which the top five are measured is demonstrated in Table 3. As can be seen from Table 3, reducing the number of genes analyzed to the 5 most predictive has little impact on the overall predictions of contagiousness. [00190] Table 1 :

[00191] Table 2: Frequency of "predictive" variables

[00193] Table 3:

All analytes:

[00194] References for genes

[00195] MAP2K6 (SEQ ID NO: 1)

AUTHORS Han J, Lee JD, Jiang Y, Li Z, Feng L and Ulevitch RJ.

TITLE Characterization of the structure and function of a novel MAP kinase (MKK6)

JOURNAL J. Biol. Chem. 271 (6), 2886-2891 (1996)

PUBMED 8621675 [00196] ATF3 (SEQ ID NO: 2)

AUTHORS Chen BP, Liang G, Whelan J and Hai T.

TITLE ATF3 and ATF3 delta Zip. Transcriptional repression versus activation by alternatively spliced isoforms

JOURNAL J. Biol. Chem. 269 (22), 15819-15826 (1994)

PUBMED 7515060 [00197] BCL2L14 (SEQ ID NO: 3)

AUTHORS Guo B, Godzik A and Reed JC.

TITLE Bcl-G, a novel pro-apoptotic member of the Bcl-2 family

JOURNAL J. Biol. Chem. 276 (4), 2780-2785 (2001)

PUBMED 11054413 [00198] CCL2 (SEQ ID NO: 4)

AUTHORS Yoshimura T and Leonard EJ.

TITLE Human monocyte chemoattractant protein- 1 (MCP-1)

JOURNAL Adv. Exp. Med. Biol. 305, 47-56 (1991)

PUBMED 1661560 [00199] CCL8 (SEQ ID NO: 5)

AUTHORS Van Damme J, Proost P, Lenaerts JP and Opdenakker G. TITLE Structural and functional identification of two human, tumor-derived monocyte chemotactic proteins (MCP-2 and MCP-3) belonging to the chemokine family

JOURNAL J. Exp. Med. 176 (1), 59-65 (1992)

PUBMED 1613466

[00200] CXCL10 (SEQ ID NO: 6)

AUTHORS Luster,A.D., Unkeless,J.C. and Ravetch,J.V.

TITLE Gamma-interferon transcriptionally regulates an early-response gene containing homology to platelet proteins

JOURNAL Nature 315 (6021), 672-676 (1985)

PUBMED 3925348

[00201] DDX58 (SEQ ID NO: 7)

AUTHORS Imaizumi T, Aratani S, Nakajima T, Carlson M, Matsumiya T, Tanji K, Ookawa K, Yoshida H, Tsuchida S, Mclntyre TM, Prescott SM, Zimmerman GA and Satoh K.

TITLE Retinoic acid-inducible gene-I is induced in endothelial cells by LPS and regulates expression of COX-2

JOURNAL Biochem. Biophys. Res. Commun. 292 (1), 274-279 (2002) PUBMED 11890704 [00202] GBP1 (SEQ ID NO: 8)

AUTHORS Cheng YS, Patterson CE and Staeheli P.

TITLE Interferon-induced guanylate -binding proteins lack an N(T)KXD consensus motif and bind GMP in addition to GDP and GTP

JOURNAL Mol. Cell. Biol. 11 (9), 4717-4725 (1991)

PUBMED 1715024

[00203] HERC6 (SEQ ID NO: 9)

AUTHORS Ebstein F, Lange N, Urban S, Seifert U, Kruger E and Kloetzel PM.

TITLE Maturation of human dendritic cells is accompanied by functional remodelling of the ubiquitin-proteasome system JOURNAL Int. J. Biochem. Cell Biol. 41 (5), 1205-1215 (2009)

PUBMED 19028597 [00204] IFIH1 (SEQ ID NO: 10)

AUTHORS Chaudhary PM, Eby MT, Jasmin A, Kumar A, Liu L and Hood L.

TITLE Activation of the NF-kappaB pathway by caspase 8 and its homologs

JOURNAL Oncogene 19 (39), 4451-4460 (2000)

PUBMED 11002417 [00205] IFIT2 (SEQ ID NO: 11)

AUTHORS Ulker N, Zhang X and Samuel CE.

TITLE Mechanism of interferon action. I. Characterization of a 54-kDa protein induced by gamma interferon with properties similar to a cytoskeletal component

JOURNAL J. Biol. Chem. 262 (35), 16798-16803 (1987)

PUBMED 3119591

[00206] LAMP3 (SEQ ID NO: 12)

AUTHORS Ozaki K, Nagata M, Suzuki M, Fujiwara T, Ueda K, Miyoshi Y, Takahashi E and Nakamura Y.

TITLE Isolation and characterization of a novel human lung-specific gene homologous to lysosomal membrane glycoproteins 1 and 2: significantly increased expression in cancers of various tissues

JOURNAL Cancer Res. 58 (16), 3499-3503 (1998) PUBMED 9721848 [00207] OAS2 (SEQ ID NO: 13)

AUTHORS Silverman RH and Sengupta DN.

TITLE Translational regulation by HIV leader RNA, TAT, and interferon-inducible enzymes JOURNAL J. Exp. Pathol. 5 (2), 69-77 (1990) PUBMED 1708818 PANK2 (SEQ ID NO: 14) AUTHORS Robishaw,J.D. and Neely,J.R.

TITLE Coenzyme A metabolism

JOURNAL Am. J. Physiol. 248 (1 PT 1), E1-E9 (1985)

PUBMED 2981478

[00208] PML (SEQ ID NO: 15)

AUTHORS Kakizuka A, Miller WH Jr, Umesono K, Warrell RP Jr, Frankel SR, Murty W, Dmitrovsky E and Evans RM.

TITLE Chromosomal translocation t(l 5; 17) in human acute promyelocytic leukemia fuses RAR alpha with a novel putative transcription factor, PML

JOURNAL Cell 66 (4), 663-674 (1991)

PUBMED 1652368

[00209] PNPT1 (SEQ ID NO: 16)

AUTHORS Raijmakers R, Egberts WV, van Venrooij WJ and Pruijn GJ.

TITLE Protein-protein interactions between human exosome components support the assembly of RNase PH-type subunits into a six-membered PNPase-like ring

JOURNAL J. Mol. Biol. 323 (4), 653-663 (2002)

PUBMED 12419256

[00210] TDRD7 (SEQ ID NO: 17)

AUTHORS Hirose T, Kawabuchi M, Tamaru T, Okumura N, Nagai K and Okada M.

TITLE Identification of tudor repeat associator with PCTAIRE 2 (Trap). A novel protein that interacts with the N-terminal domain of PCTAIRE 2 in rat brain

JOURNAL Eur. J. Biochem. 267 (7), 2113-2121 (2000)

PUBMED 10727952

[00211] TOR1B (SEQ ID NO: 18)

AUTHORS Ozelius LJ, Page CE, Klein C, Hewett JW, Mineta M, Leung J, Shalish C, Bressman SB, de Leon D, Brin MF, Fahn S, Corey DP and Breakefield XO.

TITLE The TORI A (DYT1) gene family and its role in early onset torsion dystonia JOURNAL Genomics 62 (3), 377-384 (1999)

PUBMED 10644435 SP18 (SEQ ID NO: 19)

AUTHORS Li XL, Blackford JA, Judge CS, Liu M, Xiao W, Kalvakolanu DV and Hassel BA.

TITLE RNase-L-dependent destabilization of interferon-induced mRNAs. A role for the 2- 5A system in attenuation of the interferon response

JOURNAL J. Biol. Chem. 275 (12), 8880-8888 (2000)

PUBMED 10722734

Claims

CLAIMS 1. A method for diagnosing whether a subject is contagious with an RNA respiratory virus or predicting whether a subject will become contagious with an RNA respiratory virus, the method comprising measuring the expression levels of one or more genes selected from USP18,

MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCLIO, DDX58, GBPIPI, HERC6, IFIHI, IFIT2, LAMP3, OAS2, PANK2, PML, PNPTl, TDRD7, and TORIB in a biological sample taken from the subject and assigning the subject to a class corresponding to the subject's degree or predicted degree of contagiousness by analysing the expression levels of the one or more genes and using a classification algorithm.

2. A method according to claim 1 , wherein the classification algorithm comprises a machine- learning derived algorithm derived prior to the steps of the claimed method by analysing measured expression levels of the one or more genes, measured at an early stage following inoculation with an RNA respiratory virus, in a population or sub-population of subjects who have been grouped in two or more classes according to their degree of contagiousness, as determined by their symptoms scores and viral loads.

3. A method according to any one of the preceding claims, wherein the expression levels show a pattern of increased expression of at least one of the genes, preferably two or more genes, relative to respective reference expression levels.

4. A method according to claim 3, wherein the reference expression levels of the one or more genes is at least one previously measured expression level for the gene in the same subject.

5. A method according to claim 4, wherein the reference expression level for the one or more genes comprises a baseline expression level of the gene for the subject that was measured at a time when the subject was known not to be infected with an RNA respiratory virus.

6. A method according to any one of the preceding claims, wherein the classification algorithm is derived based on subjects that are classified using a clustering algorithm based on their scores for the symptoms associated with contagiousness and viral loads, wherein the symptoms associated with contagiousness are coughing, sneezing and runny nose, following inoculation with the virus.

7. A method according to any one of the preceding claims, wherein expression levels for the one or more genes are measured up to 120 hours after possible exposure to RNA respiratory virus, or up to 60 hours after possible exposure to RNA respiratory virus, for example between 24-60 hours, or 36-60 hours after possible exposure to RNA respiratory virus, typically around 44-52 hours, in particular 40-48 hours e.g., about 48 hours or 43 hours.

8. A method according to any one of the preceding claims, wherein the expression levels of the one or more genes are measured repeatedly over a period of a few days or weeks to monitor for changes in the expression levels.

9. A method according to claim 8, wherein the expression levels of the one or more genes are monitored every 1-2 days for at least 2 days and preferably longer, e.g., 3-10 days.

10. A method according to claims 8 or 9, wherein the expression levels of the one or more genes are monitored three times a day, or twice a day, or once a day.

11. A method according to any one of the preceding claims, wherein the expression level of one or more of USP18, CXCLIO, IFIT2, ATF3, or GBPIP is measured, for example USP18, CXCLIO and IFIT2.

12. A method according to any one of the preceding claims, wherein the expression level of one or more of USP18, CXCLIO, IFIT2, ATF3, GBPIP, or CCL8, or optionally all of USP18, CXCLIO, IFIT2, ATF3, GBPIP, or CCL8, is measured.

13. A method according to any one of the preceding claims, wherein the expression level of one or more of USP18, CXCLIO, IFIT2, ATF3, GBPIP, CCL8, CCL2, or DDX58 is measured.

14. A method according to any one of the preceding claims, wherein the expression level of all of USP18, CXCLIO, IFIT2, ATF3, GBPIP, CCL8, CCL2, or DDX58 is measured.

15. A method according to any one of the preceding claims, wherein the expression level of one or more of MAP2K6, ATF3, CXCLIO, TDRD7, DDX58 or GBPIPI, preferably MAP2K6 or TDRD7, is measured.

16. A method according to any one of the preceding claims, wherein the expression level of two, three, four, five, six, seven, eight, nine, ten, eleven, twelve, thirteen, fourteen, fifteen, sixteen, seventeen, eighteen, or all nineteen of USP18, MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCLIO, DDX58, GBPIPI, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML, PNPT1, TDRD7, and TOR1B is measured.

17. A method according to any one of the preceding claims, wherein the subjects are classified into two, preferably three or more, classes according to their level of contagiousness.

18. A method according to any one of the preceding claims, wherein the classification algorithm comprises a threshold level for the one or more genes.

19. A method according to any one of the preceding claims, wherein the expression levels of two or three or more of the genes are analysed using the classification algorithm to detect a pattern of increased expression of the genes.

20. A method according to any one of the preceding claims, wherein the expression levels of a panel of two or more, or three or more, of the genes are measured.

21. A method according to claim 20, comprising measuring the expression level of a panel of genes comprising MAP2K6 or TDRD7 and one or more additional genes.

22. A method according to claim 21, wherein the panel comprises MAP2K6 and ATF3,

MAP2K6 and CXCLIO, MAP2K6 and TDRD7, MAP2K6 and DDX58, MAP2K6 and GBPIPI, ATF3 and CXCLIO, ATF3 and TDRD7, ATF3 and DDX58, ATF3 and GBPIPI, CXCLIO and TDRD7, CXCLIO and DDX58, CXCLIO and GBPIPI, TDRD7 and DDX58, TDRD7 and GBPIPI or DDX58 and GBPIPI .

23. A method according to claim 21, wherein the panel comprises MAP2K6, ATF3 and

CXCLIO; MAP2K6, ATF3 and TDRD7; MAP2K6, ATF3 and DDX58; MAP2K6, ATF3 and GBPIPI ; MAP2K6, CXCLIO and TDRD7; MAP2K6, CXCLIO and DDX58; MAP2K6, CXCLIO and GBPIPI ; MAP2K6, TDRD7 and DDX58; MAP2K6, TDRD7 and GBPIPI ;

MAP2K6, DDX58 and GBPIPI ; ATF3, CXCLIO and TDRD7; ATF3, CXCLIO and DDX58; ATF3, CXCLIO and GBPIPI ; ATF3, TDRD7 and DDX58; ATF3, TDRD7 and GBPIPI ; ATF3, DDX58 and GBPIPI ; CXCLIO, TDRD7 and DDX58; CXCLIO, TDRD7 and GBPIPI ;

CXCLIO, DDX58 and GBPIPI ; or TDRD7, DDX58 and GBPIPI .

24. A method according to 21, wherein the panel comprises MAP2K6, ATF3, CXCLIO and TDRD7; MAP2K6, ATF3, CXCLIO and DDX58; MAP2K6, ATF3, CXCLIO and GBPIPI ; MAP2K6, ATF3, TDRD7 and DDX58; MAP2K6, ATF3, TDRD7 and GBPIPI ; MAP2K6, ATF3, DDX58 and GBPIPI ; MAP2K6, CXCLIO, TDRD7 and DDX58; MAP2K6, CXCLIO, TDRD7 and GBPIPI; MAP2K6, CXCLIO, DDX58 and GBPIPI ; MAP2K6, TDRD7, DDX58 and GBPIPI ; ATF3, CXCLIO, TDRD7 and DDX58; ATF3, CXCLIO, TDRD7 and GBPIPI ; ATF3, CXCLIO, DDX58 and GBPIPI ; ATF3, TDRD7, DDX58 and GBPIPI ; or CXCLIO, TDRD7, DDX58 and GBPIPI .

25. A method according to claim 21, wherein the panel comprises MAP2K6, ATF3, CXCLIO, TDRD7, DDX58 and GBPIPI .

26. A method according to claim 21, wherein the panel comprises MAP2K6 and a least one of ATF3, CXCLIO, TDRD7, DDX58 and GBP1P.

27. A method according to claim 26, wherein the panel comprises MAP2K6; ATF3 and

CXCLIO, ATF3 and TDRD7, ATF3 and DDX58, ATF3 and GBPIPI, CXCLIO and TDRD7, CXCLIO and DDX58, CXCLIO and GBPIPI, TDRD7 and DDX58, TDRD7 and GBPIPI or

DDX58 and GBPIPI ; and optionally one or more additional genes.

28. A method according to claim 21, wherein the panel comprises TDRD7 and a least one

additional gene.

29. A method according to claim 28, wherein the panel comprises TDRD7 and a least one of MAP2K6, ATF3, CXCLIO, DDX58 and GBP IP.

30. A method according to claim 28, wherein the panel comprises TDRD7; ATF3 and CXCLIO, ATF3 and DDX58, ATF3 and GBPIPI, ATF3 and MAP2K6, CXCLIO and DDX58, CXCLIO and GBPIPI, CXCLIO and MAP2K6, DDX58 and GBPIPI, DDX58 and MAP2K6 or GBPIPI and MAP2K6; and optionally one or more additional genes.

31. A method as claimed in any preceding claim, comprising measuring the expression levels of a panel of genes comprising MAP2K6 and TDRD7 and optionally one or more additional genes.

32. A method according to claim 31 , wherein the panel comprises MAP2K6 and TDRD7; ATF3 and CXCLIO, ATF3 and DDX58, ATF3 and GBPIPI, CXCLIO and DDX58, CXCLIO and GBPIPI or DDX58 and GBPIPI ; and optionally one or more additional genes.

33. A method for controlling the spread of infection with an RNA respiratory virus such, for example, as influenza, which comprises measuring the expression levels of one or more genes selected from MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCLIO, DDX58, GBPIPI, HERC6, IFIHI, IFIT2, LAMP3, OAS2, PANK2, PML, PNPTl, TDRD7, TORIB and USP18 in biological samples taken from each member of a group of subjects; and assigning the, or each, subject to a class corresponding to the subjects' degree or predicted degree of contagiousness by analysing the expression levels of the one or more genes using a classification algorithm, and flagging the subjects identified as contagious for treatment or separation from the remainder of the group.

34. A method according to claim 33, wherein the classification algorithm comprises a machine- learning derived algorithm derived prior to the steps of the claimed method by analysing measured expression levels of the one or more genes, measured at an early stage following inoculation with an RNA respiratory virus, in a population or sub-population of subjects who have been grouped in two or more classes according to their degree of contagiousness, as determined by their symptoms scores and viral loads.

35. A method according to claim 33 or 34, wherein a reference level of expression of the one or more genes for each individual member of the group is used.

36. A method according to claim 35, comprising comparing the expression level of the one or more genes for each member of the group with a previously determined baseline expression level of the gene for the same member.

37. A method according to any one of claims 33 to 36, wherein separation from the remainder of the group includes being given a mask, or any other means employed to limit the spread of the RNA respiratory virus and/or secondary infection.

38. A method according to any one claims 33 to 37, wherein the expression level of one or more of USP18, CXCLIO, IFIT2, ATF3, or GBPIP is measured, for example USP18, CXCLIO and IFIT2.

39. A method according to any one claims 33 to 38, wherein the expression level of one or more of USP18, CXCLIO, IFIT2, ATF3, GBPIP, or CCL8, or optionally all of USP18, CXCLIO, IFIT2, ATF3, GBPIP, or CCL8, is measured.

40. A method according to any one of claims 33 to 39, wherein the expression level of one or more of USP18, CXCLIO, IFIT2, ATF3, GBPIP, CCL8, CCL2, or DDX58 is measured.

41. A method according to any one of claims 33 to 40, wherein the expression level of all of USP18, CXCLIO, IFIT2, ATF3, GBPIP, CCL8, CCL2, or DDX58 is measured.

42. A method according to any one of claims 33 to 41, wherein two, three, four, five, six, seven, eight, nine, ten, eleven, twelve, thirteen, fourteen, fifteen, sixteen, seventeen, eighteen, or all nineteen of USP18, MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCLIO, DDX58, GBP1P1, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML, PNPT1, TDRD7, and TOR1B are selected.

43. A method for controlling the spread of an RNA respiratory virus comprising measuring the expression levels of one or more genes selected from USP18, MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCLIO, DDX58, GBP1P1, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML, PNPT1 , TDRD7, and USP18 in a biological sample taken from each member of a group of subjects; transmitting data encoding the expression levels for each subject to a server for analysis and assigning the subject to a class corresponding to the subject's degree or predicted degree of contagiousness by analysing the expression levels of the one or more genes using a classification algorithm; receiving data from the remote server identifying the members of the group who are contagious or likely to become contagious and separating the flagged subjects from the remainder of the group.

44. A method according to claim 43, wherein the classification algorithm comprises a machine- learning derived algorithm derived prior to the steps of the claimed method by analysing measured expression levels of the one or more genes, measured at an early stage following inoculation with an RNA respiratory virus, in a population or sub-population of subjects who have been grouped in two or more classes according to their degree of contagiousness, as determined by their symptoms scores and viral loads.

45. A method according to any one of claims 43 to 44, wherein the expression levels of the one or more genes are analysed using the classification algorithm to detect a pattern of increased expression of the one or more genes.

46. A method according to claim 45, wherein the increased expression levels of the one or more genes relative to a respective reference level indicates a likelihood of becoming contagious for the RNA respiratory virus.

47. A method according to any one of preceding claims, further comprising treating a subject who is diagnosed as contagious with an RNA respiratory virus, or predicted to become contagious with such a virus, with an anti-viral agent.

48. A method according to claim 47, wherein said anti-viral agent is selected from amantadine, rimantadine, oseltamivir and zanamivir.

49. A method according to any one of the preceding claims, further comprising treating a

subject who is predicted to become contagious with an RNA respiratory virus with one or more medicaments for alleviating symptoms of infection, particularly sneezing, runny nose, or coughing.

50. A method according to claim 49, wherein said one or more medicaments are selected from saline solutions administered in the form of nose sprays or mist, topical nasal decongestants and oral nasal decongestants (e.g. phenylephrine and pseudoephedrine).

51. A method according to claim 49, wherein said one or more medicaments are selected from oral cough suppressants (e.g. codeine, hydrocodone, dextromethorphan and diphenhydramine), oral expectorants (e.g. guaifenesin) and/or an antitussive, and topical medicines such as camphor and menthol.

52. A method according to any one of the preceding claims, further comprising treating a subject who is predicted to become contagious with an RNA respiratory virus with an antibiotic treatment for preventing the spread of secondary bacterial infection.

53. A method according to claim 52, wherein said antibiotic treatment is selected from

penicillins, tetracyclines, cephalosporins, quinolones, lincomycins, macrolides, sulphonamides, glycopeptides, aminoglycosides and carbapenems.

54. A method according to any one of the preceding claims, wherein the RNA respiratory virus is respiratory syncytial virus, influenza virus, parainfluenza virus, metapneumovirus, rhinovirus or coronavirus.

55. A method according to any one of claims 33 to 54, wherein separation from the remainder of the group includes being given a mask, or any other means employed to limit the spread of the RNA respiratory virus and/or secondary infection.

56. A method according to any one of the preceding claims, wherein the RNA respiratory virus is influenza.

57. A method according to any one of the preceding claims, wherein the expression level of the one or more genes is measured by quantifying mRNA transcripts of the one or more genes in the biological sample.

58. A method according to claim 57, wherein a PCR-based method is used for quantifying the mRNA transcripts of the one or more genes in a biological sample.

59. A method according to claim 57, wherein a gene expression microarray is used for

quantifying the mRNA transcripts in the biological sample.

60. A method according to any one of the preceding claims, wherein biological sample is a blood sample from the subject.

61. A networked apparatus for determining whether a subject is at risk of becoming contagious with an RNA respiratory virus such, for example, as influenza, the apparatus comprising:

(a) gene expression measuring equipment that is operable to measure the expression levels of one or more genes in a biological sample taken from the subject selected from USP18, MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCLIO, DDX58, GBPIPI, HERC6, IFIHI, IFIT2, LAMP3, OAS2, PANK2, PML, PNPT1, TDRD7, and TOR1B and to encode the measured expression levels as expression level data associated with identity data identifying the subject; (b) a server that is operable to receive the expression level data and identity data and to execute program code that analyses the expression levels of the one of more genes and assign the subject to a class corresponding to the subject's degree or predicted degree of contagiousness by analysing the expression levels of the one or more genes using a classification algorithm; and

(c) one or more electronic communication components for connecting the measuring equipment to the server to allow the transmission of the expression level and identity data from the measuring device to the analyser.

62. A networked apparatus according to claim 61, wherein the classification algorithm

comprises a machine-learning derived algorithm derived prior to the steps of the claimed method by analysing measured expression levels of the one or more genes, measured at an early stage following inoculation with an RNA respiratory virus, in a population or sub-population of subjects who have been grouped in two or more classes according to their degree of

contagiousness, as determined by their symptoms scores and viral loads.

63. A classification algorithm for assigning a subject to a class corresponding to the subject's degree or predicted degree of contagiousness, wherein the classification algorithm is derived by analysing measured expression levels of one or more genes, measured at an early stage following inoculation with an RNA respiratory virus, in a population or sub-population of subjects who have been grouped in two or more classes according to their degree of contagiousness, as determined by their symptoms scores and viral loads.

64. A classification algorithm according to claim 63, wherein the one or more genes are

selected from USP18, MAP2K6, ATF3, BCL2L14, CCL2, CCL8, CXCL10, DDX58, GBP1P1, HERC6, IFIH1, IFIT2, LAMP3, OAS2, PANK2, PML, PNPT1, TDRD7, and TOR1B.

65. A classification algorithm according to claim 63 or 64, wherein the classification algorithm is derived based on subjects that are classified using a clustering algorithm based on their scores for the symptoms associated with contagiousness and viral loads, wherein the symptoms associated with contagiousness are coughing, sneezing and runny nose, following inoculation with the RNA respiratory virus.

66. A classification algorithm according to any one of claims 63 to 65, wherein the

classification algorithm is computer- implemented and comprises receiving in a computer a data set comprising expression levels of one or more genes from one or more subjects and executing on the computer software to classify the one or more subjects according to their degree, or predicted degree, of contagiousness.

67. A classification algorithm according to claim 66, wherein the classification algorithm comprises a Na^'ive Bayes classifier, a support vector machine classifier or a decision tree, or preferably a random forest classifier.

68. A computer-readable medium and/or computer program comprising instructions which, when executed by a computer, cause the computer to carry out the classification algorithm according to any one of claims 65 to 67.