PROGNOSTIC PATHWAYS FOR VIRAL INFECTIONS

Abstract

The invention relates to a method for determining whether a subject with an infection has a viral infection. The invention further relates to method for determining the cellular immune response to a viral infection or a vaccine. The methods may be performed on a blood sample obtained from a subject, and is based on the finding that specific cellular signaling pathways are active. The invention further relates to components for performing the methods and use of those components in a method of diagnosis.

Claims

1. A method for distinguishing between a bacterial and a viral infection in a blood sample obtained from a subject with an infection, based on the determined expression levels of three or more target genes of the JAK-STAT1/2 cellular signaling pathway, the method comprising: receiving the determined expression levels of the three or more target genes of the JAK-STAT1/2 cellular signaling pathway; determining the JAK-STAT1/2 cellular signaling pathway activity based on evaluating a calibrated mathematical pathway model relating the expression levels of the three or more JAK-STAT1/2 target genes to an activity level of the family of JAK-STAT1/2 transcription factor (TF) elements, the family of JAK-STAT1/2 TF elements controlling transcription of the three or more JAK-STAT1/2 target genes, the activity of the JAK-STAT1/2 cellular signaling pathway being defined by the activity level of the family of JAK-STAT1/2 TF elements, the calibrated mathematical pathway model being a model that is calibrated using a ground truth dataset including samples in which transcription of the three or more JAK-STAT1/2 target genes is induced by the family of JAK-STAT1/2 TF elements and samples in which transcription of the three or more JAK-STAT1/2 target genes is not induced by the family of JAK-STAT1/2 TF elements; wherein the JAK-STAT1/2 cellular signaling pathway refers to a signaling process that leads to transcriptional activity of the family of JAK-STAT1/2 TF elements, and wherein the family of JAK-STAT1/2 TF elements are protein complexes each containing a homodimer or a heterodimer comprising STAT1 and/or STAT2; and wherein the determining the JAK-STAT1/2 cellular signaling pathway activity comprises assigning a numeric value to the JAK-STAT1/2 cellular signaling pathway activity level by evaluating the calibrated mathematical pathway model relating expression levels of the target genes to the activity level of the JAK-STAT1/2 cellular signaling pathway; and comparing the JAK-STAT1/2 cellular signaling pathway activity determined in the blood sample obtained from the subject with the JAK-STAT1/2 cellular signaling pathway activity determined in a reference blood sample, wherein the reference blood sample is obtained from a healthy subject or a subject recovered from an infection; and wherein the infection in the subject from which the blood sample is obtained is determined to be viral when the JAK-STAT1/2 cellular signaling pathway activity is higher compared to the JAK-STAT1/2 cellular signaling pathway activity in the reference blood sample, or wherein the infection in the subject from which the blood sample is obtained is determined to be bacterial when the JAK-STAT1/2 cellular signaling pathway activity is not higher compared to the JAK-STAT1/2 cellular signaling pathway activity in the reference blood sample.

2. A method for determining the cellular immune response to a viral infection or a vaccine in a blood sample obtained from a subject with a viral infection or a subject who received a vaccine, based on the determined expression levels of three or more target genes of the JAK-STAT1/2 cellular signaling pathway, the method comprising: receiving the determined expression levels of the three or more target genes of the JAK-STAT1/2 cellular signaling pathway; determining the JAK-STAT1/2 cellular signaling pathway activity based on evaluating a calibrated mathematical pathway model relating the expression levels of the three or more JAK-STAT1/2 target genes to an activity level of the family of JAK-STAT1/2 transcription factor (TF) elements, the family of JAK-STAT1/2 TF elements controlling transcription of the three or more JAK-STAT1/2 target genes, the activity of the JAK-STAT1/2 cellular signaling pathway being defined by the activity level of the family of JAK-STAT1/2 TF elements, the calibrated mathematical pathway model being a model that is calibrated using a ground truth dataset including samples in which transcription of the three or more JAK-STAT1/2 target genes is induced by the family of JAK-STAT1/2 TF elements and samples in which transcription of the three or more JAK-STAT1/2 target genes is not induced by the family of JAK-STAT1/2 TF elements, wherein the JAK-STAT1/2 cellular signaling pathway refers to a signaling process that leads to transcriptional activity of the family of JAK-STAT1/2 TF elements, and wherein the family of JAK-STAT1/2 TF elements are protein complexes each containing a homodimer or a heterodimer comprising STAT1 and/or STAT2, wherein the determining the JAK-STAT1/2 cellular signaling pathway activity comprises assigning a numeric value to the JAK-STAT1/2 cellular signaling pathway activity level by evaluating a calibrated mathematical pathway model relating expression levels of the target genes to the activity level of the JAK-STAT1/2 cellular signaling pathway; and comparing the JAK-STAT1/2 cellular signaling pathway activity determined in the blood sample obtained from the subject with a viral infection or a subject who received a vaccine with the JAK-STAT1/2 cellular signaling pathway activity determined in a reference blood sample obtained from a healthy subject; wherein the activity of the JAK-STAT1/2 cellular signaling pathway is compared with the cellular signaling pathway activities determined in the reference blood samples in order to determine whether the immune response to the viral infection is weak or strong.

3. The method of claim 2, wherein the method further comprises: receiving the determined expression levels of the three or more target genes of the JAK-STAT3 cellular signaling pathway; determining the JAK-STAT3 cellular signaling pathway activity, wherein the determining the JAK-STAT3 cellular signaling pathway activity comprises assigning a numeric value to the JAK-STAT3 cellular signaling pathway activity level by evaluating a calibrated mathematical pathway model relating expression levels of the target genes to the activity level of the JAK-STAT3 cellular signaling pathway; and comparing the JAK-STAT3 cellular signaling pathway activity determined in the blood sample obtained from the subject with a viral infection or a subject who received a vaccine with the JAK-STAT3 cellular signaling pathway activity determined in a reference blood sample obtained from a healthy subject.

4. The method of claim 2, wherein the immune response to a viral infection is considered weak when the numeric value assigned to the JAK-STAT1/2 cellular signaling pathway activity in the blood sample obtained from the subject with a viral infection or a subject who received a vaccine is one standard deviation higher than the numerical value assigned to the JAK-STAT1/2 cellular signaling pathway activity in the reference blood sample of the healthy subject and the immune response to a viral infection is considered strong when the numeric value assigned to the JAK-STAT1/2 cellular signaling pathway activity in the blood sample obtained from the subject with a viral infection or a subject who received a vaccine is at least two, preferably three or more, standard deviations higher than the numerical value assigned to the JAK-STAT1/2 cellular signaling pathway activity in the reference blood sample of the healthy subject.

5. The method of claim 2, wherein comparing the JAK-STAT1/2 and optionally the JAK-STAT3 cellular signaling pathway activities determined in the blood sample obtained from the subject with a viral infection or the subject who received a vaccine further comprises comparing with the JAK-STAT1/2 and optionally the JAK-STAT3 cellular signaling pathway activities determined in a reference blood sample obtained from a reference patient with a weak immune response and the JAK-STAT1/2 and optionally the JAK-STAT3 cellular signaling pathway activities determined in a reference blood sample obtained from a reference patient with a strong immune response, and wherein the strength of the immune response in the subject with a viral infection or the subject who received a vaccine is based on the comparison between the JAK-STAT1/2 cellular signaling pathway activities from the subject with a viral infection or the subject who received a vaccine with the JAK-STAT1/2 cellular signaling pathway activities determined in the reference blood samples obtained from the reference patient with a weak immune response and the reference blood samples obtained from the reference patient with a strong immune response.

6. The method of claim 2, wherein the JAK-STAT1/2 cellular signaling pathway activity corresponds to the strength of the immune response, wherein a higher JAK-STAT1/2 cellular signaling pathway activity signifies a stronger immune response.

7. The method of claim 3, wherein the blood sample is from a subject with a viral infection and wherein a higher JAK-STAT3 cellular signaling pathway activity is indicative of a more severe infection.

8. The method of claim 3, wherein the blood sample is from a subject who receives a vaccine and wherein a stronger immune response and optionally a higher JAK-STAT3 cellular signaling pathway activity is indicative of a stronger cellular immunity.

9. The method of claim 1, wherein the determined activity levels of the JAK-STAT1/2 and optionally the JAK-STAT3 cellular signaling pathways are further used to: monitor a patient with an infection; or determine the strength of the cellular immunity induced by a viral infection or vaccination in an individual; or predict the response to an immune modulatory therapy or drug; or monitor the response to a drug or therapy; or predict the toxicity of an immunomodulatory therapy or drug; or estimate the strength of the cellular immunity that will result in a community during an viral infection epidemic/pandemic; or determine the strength of the immunity induced by viral infection or vaccination in an individual with a specific immune compromising condition, such as a specific comorbidity, therapy, lifestyle; or diagnose patients with an viral infection during an epidemic or pandemic; or develop an drug or therapy to treat the infectious disease; or predict the immune activating or immune suppressive state caused by the viral infection.

10. The method of claim 1, wherein the method further comprises the step of determining the expression levels of the three or more target genes of the JAK-STAT1/2 cellular signaling pathway and optionally the three or more target genes of the JAK-STAT3 cellular signaling pathway and/or further comprises the step of providing or obtaining the blood sample from the subject.

11. The method of claim 1, wherein the blood sample is whole blood sample, a peripheral blood mononuclear cell sample, or isolated blood cells such as dendritic cells, CD4+ T cells, CD8+ T cells, CD16− monocytes, CD16+ monocytes, Neutrophils, NK cells and B cells.

12. The method of claim 1, wherein the three or more target genes of the JAK-STAT1/2 cellular signaling pathway are selected from the group consisting of: BID, GNAZ, IRF1, IRF7, IRF8, IRF9, LGALS1, NCF4, NFAM1, OAS1, PDCD1, RAB36, RBX1, RFPL3, SAMM50, SMARCB1, SSTR3, ST13, STAT1, TRMT1, UFD1L, USP18, ZNRF3, GBP1, TAP1, ISG15, APOL1, IFI6, IFIRM1, CXCL9, APOL2, IFIT2 and LY6E, preferably, from the group consisting of: IRF1, IRF7, IRF8, IRF9, OAS1, PDCD1, ST13, STAT1 and USP1 or from the group consisting of GBP1, IRF9, STAT1, TAP1, ISG15, APOL1, IRF1, IRF7, IFI6, IFIRM1, USP18, CXCL9, OAS1, APOL2, IFIT2 and LY6E, and/or wherein the three or more target genes of the JAK-STAT3 cellular signaling pathway are selected from the group consisting of: AKT1, BCL2, BCL2L1, BIRC5, CCND1, CD274, CDKNIA, CRP, FGF2, FOS, FSCN1, FSCN2, FSCN3, HIFIA, HSP90AA1, HSP90AB1, HSP90B1, HSPA1A, HSPA1B, ICAM1, IFNG, IL10, JunB, MCL1, MMP1, MMP3, MMP9, MUC1, MYC, NOS2, POU2F1, PTGS2, SAA1, STAT1, TIMP1, TNFRSF1B, TWIST1, VIM and ZEB1.

13. The method of claim 1, wherein the activity of the JAK-STAT1/2 cellular signaling pathway and optionally the JAK-STAT3 cellular signaling pathway in the blood sample is inferable by a method comprising: receiving expression levels of three or more target genes of the JAK-STAT1/2 cellular signaling pathway and optionally the JAK-STAT3 cellular signaling pathway, determining an activity level of a signaling pathway associated transcription factor (TF) element, the signaling pathway associated TF element controlling transcription of the three or more target genes, the determining being based on evaluating a calibrated mathematical pathway model relating expression levels of the target genes to the activity level of the JAK-STAT1/2 cellular signaling pathway and optionally the JAK-STAT3 cellular signaling pathway; and inferring the activity of the JAK-STAT1/2 cellular signaling pathway and optionally the JAK-STAT3 cellular signaling pathway in the blood sample based on the determined activity level of the signaling pathway associated TF element; wherein the calibrated mathematical pathway model is preferably a centroid or a linear model, or a Bayesian network model based on conditional probabilities.

14. An apparatus comprising at least one digital processor configured to perform the method of claim 1.

15. A non-transitory storage medium storing instructions that are executable by a digital processing device to perform the method claim 1.

16. A computer program comprising program code means for causing digital processing device to perform the method claim 1, when the computer program is run on a digital processing device.

17-20. (canceled)

21. A method for stratifying a subject with a COVID-19 infection for suitability for treatment with an AR pathway inhibitor based on a blood sample obtained from the subject with a COVID-19 infection, based on the determined expression levels of three or more target genes of the AR cellular signaling pathway, the method comprising: receiving the determined expression levels of the three or more target genes of the AR cellular signaling pathway; determining the AR cellular signaling pathway activity based on evaluating a calibrated mathematical pathway model relating the expression levels of the three or more AR target genes to an activity level of the family of AR transcription factor (TF) elements, the family of AR TF elements controlling transcription of the three or more AR target genes, the activity of the AR cellular signaling pathway being defined by the activity level of the family of AR TF elements, the calibrated mathematical pathway model being a model that is calibrated using a ground truth dataset including samples in which transcription of the three or more AR target genes is induced by the family of AR TF elements and samples in which transcription of the three or more AR target genes is not induced by the family of AR TF elements, wherein the AR cellular signaling pathway refers to a signaling process that leads to transcriptional activity of the family of AR TF elements, and wherein the family of AR TF elements are protein complexes each containing a homodimer or a heterodimer comprising AR-A and/or AR-B; wherein the determining the AR cellular signaling pathway activity comprises assigning a numeric value to the AR cellular signaling pathway activity level by evaluating the calibrated mathematical pathway model relating expression levels of the target genes to the activity level of the AR cellular signaling pathway; comparing the AR cellular signaling pathway activity determined in the blood sample obtained from the subject with the AR cellular signaling pathway activity determined in a reference blood sample; wherein the reference blood sample is obtained from a healthy subject; and wherein the subject with the COVID-19 infection is not a candidate for treatment with an AR inhibitor if the determine AR pathway activity is equal or lower than the AR pathway activity determined in the reference blood sample of the healthy subject, or wherein the subject with the COVID-19 infection is a candidate if the determined AR pathway activity is higher than the AR pathway activity determined in the reference blood sample of the healthy subject.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0175] General: In all the figures where signal transduction pathway analysis scores are depicted, these are given as log2odds scores for pathway activity, derived from the probability scores for pathway activity provided by the Bayesian pathway model analysis. Log2odds scores indicate the level of activity of a signaling pathway on a linear scale.

[0176] Analyzed public datasets are indicated with their GSE number (in principle at the bottom of each figure), and individual samples with their GSM number (in principle most right column for clustering diagrams).

[0177] All validation samples for a signaling pathway model or an immune response/system model are independent samples and have not been used for calibration of the respective model to be validated.

[0178] FIG. 1 depicts a diagram showing different cells in the blood involved in the active immune response. The immune response can be divided in the innate immune response and adaptive immune response, each of which is controlled by different cell types as depicted.

[0179] FIG. 2 depicts pathway activity analysis performed on GSE6269. Peripheral blood samples from pediatric patients with acute infections of Influenza, S. pneumoniae or S. aureus were used for the pathway analysis. Only significant pathways using Mann-Whitney-Wilcoxon test are included in the figure. Both the JAK-STAT1/2 type I and type II interferon are increased in only influenza but not in the bacterial infections.

[0180] FIG. 3 depicts pathway activity analysis of dataset GSE4607 of children <10 years of age admitted to the pediatric intensive care unit and meeting the criteria for septic shock caused by a bacterial infection. Shown are AR and TGFbeta signaling activity. Significance was calculated using Mann-Whitney-Wilcoxon test and are included in the figure.

[0181] FIG. 4 depicts pathway activity analysis of dataset GSE4607 of children <10 years of age admitted to the pediatric intensive care unit and meeting the criteria for septic shock caused by a bacterial infection. Shown are JAK-STAT1/2 type I and type II interferon signaling activity. Significance was calculated using Mann-Whitney-Wilcoxon test and are included in the figure.

[0182] FIG. 5 depicts pathway analysis of dataset GSE69606 including in pediatric Respiratory Syncytial Virus (RSV) (mild, moderate, severe) and recovery (moderate and severe) patients. Analysis was performed on PBMCs. Shown are JAK-STAT1/2 type I and type II interferon signaling activity.

[0183] FIG. 6 depicts pathway analysis of dataset GSE69606 including in pediatric Respiratory Syncytial Virus (RSV) (mild, moderate, severe) and recovery (moderate and severe) patients. Analysis was performed on PBMCs. Shown are JAK-STAT3 and AR signaling activity.

[0184] FIG. 7 depicts Pathway analysis of dataset GSE43777, dengue viral infection measured in PBMC samples from DF (dengue fever) and DHF (dengue hemorrhagic fever) with acute infection. JAK-STAT1/2 type I and type II interferon signaling pathway scores including Mann-Whitney-Wilcoxon test scores are shown in the figure. Each stage represents a group of patients after x days from fever onset, Stages (days after onset fever): G1: 0 days, G2: 2 days, G3:4, G4:4 days, G5, 5 days, G6: 6-10 days, G7: >20 days. Legend and significance are shown in figure.

[0185] FIG. 8 depicts Pathway analysis of dataset GSE43777, dengue viral infection measured in PBMC samples from DF (dengue fever) and DHF (dengue hemorrhagic fever) with acute infection. JAK-STAT3 signaling pathway scores including Mann-Whitney-Wilcoxon test scores are shown in the figure. Each stage represents a group of patients after x days from fever onset, Stages (days after onset fever): G1: 0 days, G2: 2 days, G3:4, G4:4 days, G5, 5 days, G6: 6-10 days, G7: >20 days. Legend and significance are shown in figure. JAK-STAT3 pathway activity is higher in infections with a more severe clinical course.

[0186] FIG. 9 depicts pathway analysis of dataset GSE34205 including patients hospitalized with acute RSV and influenza virus infection. Blood samples were collected from children within 42-72 hours of hospitalization. The PBMCs were used for the study. The pathways STAT1/2 type I and type II interferon signaling pathway activity are shown and significance determined using Mann-Whitney-Wilcoxon test. Legend and significance are included in the figure. The strength of the generated cellular immunity by the infection is reflected by activity of the JAK-STAT1/2 pathway, and varies with type if viral infection.

[0187] FIG. 10 depicts pathway analysis of dataset GSE34205 including patients hospitalized with acute RSV and influenza virus infection. Blood samples were collected from children within 42-72 hours of hospitalization. The PBMCs were used for the study. The pathways STAT3 and TGFbeta signaling pathway activity are shown and significance determined using Mann-Whitney-Wilcoxon test. Legend and significance are included in the figure.

[0188] FIG. 11 depicts pathway analysis of dataset GSE119322, dendritic cell function in patients with Hepatitis carriers with normal liver function, untreated chronic hepatitis and chronic hepatitis patients undergoing nucleoside analogue treatment. Pathways JAK-STAT1/2 (type I and type II interferon) and JAK-STAT3 are shown including their significance using Mann-Whitney-Wilcoxon test. JAK-STAT1 pathway activity is lower in chronic hepatitis B patients, with suppressed cellular immunity.

[0189] FIG. 12 depicts pathway analysis of dataset GSE51997 comprising healthy and yellow fever (YFV-17D) vaccinated volunteers. CD4 and CD16 negative and positive cells were isolated and used in the study. Depicted are the JAK-STAT1/2 (type I and type II interferon) and the JAK-STAT3 pathway activities. Legend and significance are included in the figure.

[0190] FIG. 13 depicts pathway analysis of dataset GSE51997 comprising healthy and yellow fever (YFV-17D) vaccinated volunteers. CD4 and CD16 negative and positive cells were isolated and used in the study. Depicted are the TGFbeta and the NFkB pathway activities. Legend and significance are included in the figure. JAK-STAT pathway activity increases within days after viral vaccination.

[0191] FIG. 14 depicts Pathway analysis of dataset GSE29614. Time Course of Young Adults Vaccinated with Influenza TIV (Trivalent Inactivated Influenza Vaccine) Vaccine during the 2007/08 Flu Season. Blood samples isolated at days 0, 3, 7 days post-vaccination. Depicted are the JAK-STAT1/2 (type I and type II interferon) and the JAK-STAT3 pathway activities. Legend and significance using Mann-Whitney-Wilcoxon test are shown in the figure.

[0192] FIG. 15 depicts pathway analysis of dataset GSE22589 containing uninfected human monocyte-derived dendritic cells (MDDCs); MDDCs infected with an envelope-defective GFP-encoding VSV-G-pseudotyped HIV-1 vector (HIVGFP(G)) and with VSV-G pseudotyped virus-like particles derived from SIVmac to deliver Vpx (SIVVLP(G)), alone or in combination. Cells were infected at day 4 of differentiation and harvested 48 hours later. Depicted are the JAK-STAT1/2 (type I and type II interferon) pathway activities. Legend and significance are included in the figure. JAK-STAT1/2 pathway activity is only increased when the combined vectors are introduced into the cells and only this combination was considered as eliciting a successful response.

[0193] FIG. 16 depicts pathway analysis of dataset GSE22589 containing uninfected human monocyte-derived dendritic cells (MDDCs); MDDCs infected with an envelope-defective GFP-encoding VSV-G-pseudotyped HIV-1 vector (HIVGFP(G)) and with VSV-G pseudotyped virus-like particles derived from SIVmac to deliver Vpx (SIVVLP(G)), alone or in combination. Cells were infected at day 4 of differentiation and harvested 48 hours later. Depicted are the JAK-STAT3 and NFkB pathway activities. Legend and significance are included in the figure.

[0194] FIG. 17 depicts pathway analysis of dataset GSE35283 containing monocytes infected with low (PR8) and high pathogenic influenza viruses (FPV (H7N7) and H5N1). Depicted are the JAK-STAT1/2 (type I and type II interferon) and the JAK-STAT3 pathway activities. Legend and significance are included in the figure. JAK-STAT1/2 pathway activity distinguishes between low and high pathogenic virus.

[0195] FIG. 18 depicts pathway analysis of dataset GSE50628 of influenza A(H1N1)pdm09 and rotavirus gastroenteritis infected patients. Whole blood samples were collected from patients in the acute phase of the disease and in the recovery phase. Significance was calculated using Mann-Whitney-Wilcoxon test. Depicted are the JAK-STAT1/2 (type I and type II interferon) pathway activities. Legend and significance are shown in the figure.

[0196] FIG. 19 depicts pathway analysis of dataset GSE50628 of influenza A(H1N1)pdm09 and rotavirus gastroenteritis infected patients. Whole blood samples were collected from patients in the acute phase of the disease and in the recovery phase. Significance was calculated using Mann-Whitney-Wilcoxon test. Depicted are the TGFbeta and the JAK-STAT3 pathway activities. Legend and significance are shown in the figure.

[0197] FIG. 20 Pathway analysis of dataset GSE84331, CD8 T Cell Responses in Dengue Virus-Infected Patients from India. The HLA−DR+CD38+CD8 T cells from the PBMCs of dengue patients from Siriraj Hospital in Bangkok, Thailand and compare with the sorted naive (CCR7+CD45RA+) CD8 T cells from Thai healthy donors. Pathways JAK-STAT1/2 and 3 and Mann-Whitney-Wilcoxon test results are shown.

[0198] FIG. 21 depicts a summary of the determined pathway activities per dataset and indicating the sample type. Upwards arrows indicate increased pathway activity, downwards arrows indicate reduced pathway activity.

[0199] FIG. 22 depicts pathway analysis of tuberculosis vaccine candidate dataset GSE102459. Human subjects were injected with M72/AS01E tuberculosis vaccine candidate at D0 and D30, whole blood samples are collected at D0, D30, D31, D37, D40, D44 and D47 and PBMC samples are collected at D0, D31, D44. Pathway activity scores are shown in log2odds values. Two sided Mann-Whitney-Wilcoxon tests were performed; p-values are indicated in the figures. P-values indicate: *p<0.05; **p<0.01; ***/****p<0.001. Depicted are the JAK-STAT1/2 IFN-I and IFN II pathway activities. Legend is shown in the figure.

[0200] FIG. 23 depicts pathway analysis of tuberculosis vaccine candidate dataset GSE102459. Human subjects were injected with M72/AS01E tuberculosis vaccine candidate at D0 and D30, whole blood samples are collected at D0, D30, D31, D37, D40, D44 and D47 and PBMC samples are collected at DO, D31, D44. Pathway activity scores are shown in log2odds values. Two sided Mann-Whitney-Wilcoxon tests were performed; p-values are indicated in the figures. P-values indicate: *p<0.05; **p<0.01; ***/****p<0.001. Depicted are the NFkB and JAK-STAT3 pathway activities. Legend is shown in the figure.

[0201] FIG. 24 depicts Pathway analysis of malaria vaccines, dataset GSE89292. Human subjects were vaccinated using ARR of RRR and challenged with Malaria. Participants in both study arms were vaccinated at 28-day intervals, and subjected to controlled malaria infection 21 days following the final immunization. Pathway activity scores are shown in log2odds values. Two sided Mann-Whitney-Wilcoxon tests were performed; p-values are indicated in the figures. P-values indicate: *p<0.05; **p<0.01; ***/****p<0.001. Depicted are the JAK-STAT1/2 IFN-I and IFN II and the JAK-STAT3 pathway activities. Legend is shown in the figure.

[0202] FIG. 25 depicts pathway analysis of dataset GSE22589 containing uninfected human monocyte-derived dendritic cells (MDDCs); MDDCs infected with an envelope-defective GFP-encoding VSV-G-pseudotyped HIV-1 vector (HIVGFP(G)) and with VSV-G pseudotyped virus-like particles derived from SIVmac to deliver Vpx (SIVVLP(G)), alone or in combination. Cells were infected at day 4 of differentiation and harvested 48 hours later. Depicted is the JAK-STAT1/2 pathway activity using the 23 gene (23g), the 16 gene (16g) and 4 randomly selected 3 gene models (11g3, 12g3, 13g3 and 14g3).

[0203] FIG. 26 depicts Pathway analysis of dataset GSE29614. Time Course of Young Adults Vaccinated with Influenza TIV (Trivalent Inactivated Influenza Vaccine) Vaccine during the 2007/08 Flu Season. Blood samples isolated at days 0, 3, 7 days post-vaccination. Depicted is the JAK-STAT1/2 pathway activity using the 23 gene (23g), the 16 gene (16g) and 4 randomly selected 3 gene models (11g3, 12g3, 13g3 and 14g3).

[0204] FIG. 27 depicts pathway analysis of dataset GSE34205 including patients hospitalized with acute RSV and influenza virus infection. Blood samples were collected from children within 42-72 hours of hospitalization. The PBMCs were used for the study. Depicted is the JAK-STAT1/2 pathway activity using the 23 gene (23g), the 16 gene (16g) and 4 randomly selected 3 gene models (11g3, 12g3, 13g3 and 14g3).

[0205] FIG. 28 depicts pathway analysis of dataset GSE35283 containing monocytes infected with low (PR8) and high pathogenic influenza viruses (FPV (H7N7) and H5N1). Depicted is the JAK-STAT1/2 pathway activity using the 23 gene (23g), the 16 gene (16g) and 4 randomly selected 3 gene models (11g3, 12g3, 13g3 and 14g3).

[0206] FIG. 29 depicts Pathway analysis of dataset GSE43777, dengue viral infection measured in PBMC samples from DF (dengue fever) and DHF (dengue hemorrhagic fever) with acute infection. Each stage represents a group of patients after x days from fever onset, Stages (days after onset fever): G1: 0 days, G2: 2 days, G3:4, G4:4 days, G5, 5 days, G6: 6-10 days, G7: >20 days. Depicted is the JAK-STAT1/2 pathway activity using the 23 gene (23g), the 16 gene (16g) and 4 randomly selected 3 gene models (11g3, 12g3, 13g3 and 14g3).

[0207] FIG. 30 depicts pathway analysis of dataset GSE50628 of influenza A(H1N1)pdm09 and rotavirus gastroenteritis infected patients. Whole blood samples were collected from patients in the acute phase of the disease and in the recovery phase. Depicted is the JAK-STAT1/2 pathway activity using the 23 gene (23g), the 16 gene (16g) and 4 randomly selected 3 gene models (11g3, 12g3, 13g3 and 14g3).

[0208] FIG. 31 depicts pathway analysis of dataset GSE51997 comprising healthy and yellow fever (YFV-17D) vaccinated volunteers. CD4 and CD16 negative and positive cells were isolated and used in the study. Depicted is the JAK-STAT1/2 pathway activity using the 23 gene (23g), the 16 gene (16g) and 4 randomly selected 3 gene models (11g3, 12g3, 13g3 and 14g3).

[0209] FIG. 32 depicts pathway analysis of dataset GSE13486 comprising healthy yellow fever (YFV-17, live attenuated yellow fever virus strain without adjuvant) vaccinated volunteers. PBMCs were isolated and used in the study. Depicted are the JAK-STAT1/2 (type I and type II interferon) pathway activities. Legend and significance using Mann-Whitney-Wilcoxon test are shown in the figure.

[0210] FIG. 33 depicts RNA sequencing pathway analysis of dataset GSE157103 [12] containing COVID-19 and non COVID patients with respiratory issues. ICU and ventilation status is indicated. A) AR pathway activity scores, B) AP1 pathway activity scores and C) STAT1-2 pathway activity scores. Patient samples are shown as duplicate measurements (each sample was measured twice, and shown in the figure).

[0211] FIG. 34 depicts pseudotime (disease trajectory) scores (A) and longitudinal patient data categorized based on severity (B). see details [15].

[0212] FIG. 35 depicts AR (left), AR (middle) and STAT1-2 (right) RNA sequencing pathway activity scores of patients 1, 2, 4, 10 and HC of dataset GSE161777 [15]. containing COVID patients and Healthy controls (HC). COVID-19 patients are categorized per severity; Critical ill patients (3A), complicated patients (3B), Moderate/early convalescence (3C) and Late convalescence/Recovery (3D).

[0213] FIG. 36 depicts AR (left), AR (middle) and STAT1-2 (right) RNA sequencing pathway activity scores of patients 1, 2, 4, 10 and HC of dataset GSE161777 [15]. All COVID-19 patients showed disease worsening and improvement, except for patient 2 where disease progression lead to death.

EXAMPLES

Methods and Sample Description

[0214] Using the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/gds/) Affymetrix HG-U133Plus2.0 datasets from clinical and preclinical studies were used. Information about the used datasets, sample type and and preparations can be found in Table 1 below. The literature references related to the original datasets are also included in the table. We used the pathway analysis to determine the signal transduction pathway activities (included pathways; AR, ER, PR, GR, HH, Notch, TGFbeta, WNT, JAK-STAT1/2, JAK-STAT3, NFkB, PI3K, MAPK). We compared the different pathway activities within the groups per dataset. Results per dataset can be found in the figure description.

[0215] The pathway analysis method to measure the cellular immune response was used to analyze Affymetrix U133P2.0 expression microarray data from a number of clinical studies of patients infected with a virus, or vaccinated with a specific vaccine, and preclinical studies on vaccine development. In the pathway analysis method, a calibrated mathematical method was used for each pathway to relate the expression levels of target genes for each pathway to the pathway activity. The pathway activity is represented by a numerical value. In all viral infections that we studied to date in which patients recovered, increased activity of the JAK-STAT1/2 pathway was observed, in combination with changes in a number of other immune-modulating signaling pathways (JAK-STAT3, TGFbeta, NFkB and AR pathway) which were informative for the raised cellular immune response, where the increases in JAK-STAT3 pathway activity was especially informative on the clinical course of the infection (severity).

Distinction Between Viral and Bacterial Disease (FIGS. 2-4)

[0216] Pathway analysis from a number of clinical viral studies shows that we can distinguish on an individual patient basis between viral and bacterial infection (FIGS. 2-4). For example the JAK-STAT1/2 pathway activity was specifically increased in viral infections, not in bacterial infections. Therefore, the JAK-STAT1/2 pathway activity (type I and type II interferon) is and indicator of viral infection and not for bacterial infection in which among other AR and TGFbeta is of importance.

[0217] Both AR and TGFbeta pathways are increased in sepsis shock survivors and non survivors patients compared to healthy controls. JAK-STAT1-2 INFI and INFII pathway activity is only increased in a selection of patients. The patients with the highest expression of JAK-STAT1-2 patients (n=5) had an lung infection. From community-acquired pneumonia its known that 70% of hospitalized CAP patients initially have sepsis or may develop sepsis during their hospital stay. It is likely these sepsis patients with an high JAK-STAT1-2 activity have an co infection of bacterial and viral infection.

Distinction Between Mild and Severe Infection (FIGS. 5-8)

[0218] The pathway analysis showed that both the JAK-STAT1-2 INF 1 and II were increased in RSV infected patients compared to the control. JAK-STAT3 activity was progressively increased with the severity of the disease. Significance using Mann-Whitney-Wilcoxon test and legend are shown in the figure. We could distinguish between mild and severe infection. In pediatric patients infected with RSV (respiratory syncytial virus, which shows some clinical similarities to COVID-19) we could distinguish between mild and severe respiratory infections based on PBMC blood cell analysis (FIGS. 5 and 6). In Dengue viral infection similarly JAK-STAT3 pathway activity was higher in the more serious hemorrhagic fever variant (FIGS. 7 and 8). Both groups show a decrease in pathway activity of JAK-STAT1-2 INF I and INF II after day 4 of fever onset, which reflects the known immunosuppressive effect of the virus on the adaptive cellular immune response. There was a Higher JAK-STAT3 pathway activity in the G2 DHF group compared to the G2 DF group.

Strength of the Cellular Immune Response in an Individual and in a Population (FIGS. 9-11)

[0219] We could measure on patient PBMCs that the immune response to influenza virus infection was stronger than to RSV virus, in line with literature describing that in general in the population the immune response generated by the RSV virus is lower and less persistent (FIGS. 9 and 10). The Influenza infected patients induces a higher JAK-STAT1-2 INF I and II pathway activity compared to RSV infected patients, which is in agreement with the authors. Furthermore, JAK/STAT3 and TGFbeta pathway activity slightly increased in influenza, not in RSV. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3347264/

[0220] In PBMCs of chronic hepatitis patients the cellular immune response was minimal or absent, in line with the lack of cellular immune response in this chronic persistent infection (FIG. 11). Hepatitis B infection was associated with reduced JAK-STAT3 activity in DCs.

[0221] Similarly, The Dengue virus is known to cause suppression of the cellular immune response, which is indeed measured in the clinical study on Dengue viral infection, where patients have been monitored over the course of the disease (Day 1-6, and day 20=recovery): in contrast to what we observed in influenza, RSV, rota, and yellow fever virus infections, here the JAK_STAT1/2 pathway peaks at the beginning of the disease (Day 1), and subsequently goes down instead of up (FIGS. 7 and 8).

In general within a population with a certain type of viral infection a large range in the strength of the cellular immune response was observed, in line with common knowledge that the cellular immunity is determined by many factors that are variably present in the population, such as comorbidities, lifestyle factors, use of drugs, old age (all figures).

Measuring the Effectiveness of Vaccination in Humans (FIGS. 12-14)

[0222] In clinical vaccination studies we could distinguish between effective and ineffective vaccines (FIGS. 12-14). The significant pathways JAK-STAT1/2/3, TGFBeta and NFKB using T-test are included in the figure. The yellow fever vaccination induces JAK-STAT1-2 type I and II IFN pathway activity in all cell types, which is in agreement with the authors of the paper, also the JAK-STAT3 pathway activity was increased in the vaccinated individuals compared to the control. NFKB was only increased in the inflammatory monocytes (CD16−) and TGFBeta only in non-inflammatory monocytes (CD16+). Yellow fever vaccine in general appeared to be a highly effective vaccination, inducing a strong cellular immune response (FIGS. 12 and 13). However, although reported to produce a humoral immune response, a clinically tested influenza vaccine appeared to not induce a cellular immune response (compare FIG. 14 with FIGS. 9 and 10). This is in line with the general observation that immunity generated by influenza vaccination is not as protective against influenza infection as having recovered from an influenza infection.

[0223] There is no change in JAK-STAT1/2 and JAK-STAT3 pathway activity upon vaccination. However, the authors measured an humoral protection measured (FIG. 14). This humoral activity was not measured using the pathway analysis which only measured the cellular immune response. A combined cellular and humoral immune response provides better protection against influenza than only an humoral immune response. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3140559, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6605831/

Measuring the Effectiveness of Vaccination in Experimental Setting for Vaccine Development (FIGS. 15-17)

[0224] Vaccine development with initial effectiveness/efficiency testing requires testing in in vitro cell model systems. For example by introduction of the vaccine in monocytes, macrophages or dendritic cells to investigate how they stimulate antigen processing and presentation within HLA on the cell membrane. Using pathway analysis of sample data from vaccine-infected cells, we show that the effectiveness of inducing antigen presentation can be quantitatively measured, in accordance with the reported behavior of the vaccine. The JAK-STAT signaling pathway was clearly the indicative signaling pathway, in line with its prominent immune role in antigen presentation.

[0225] Two non-efficient HIV vaccines did not elicit an increase in activity of the JAK-STAT1/2 pathway when introduced into dendritic cells, while the combination was successful (FIGS. 15 and 16). Only a combination of combined HIVGFP(G)) and SIVVLP(G) infection resulted in an JAK-STAT1-2 response. Which is in agreement with author's in which MDDC infected with both HIVGFP(G) and SIVVLP(G) stimulated a high proportion of MHC class I and class II restricted T cell clones to produce IFNγ. Other pathways which were only significant different using the combination are the TGFBeta and NFKB pathway.

[0226] For vaccine development it is also important to compare effects of low pathogenetic (potentially to be used for vaccination purposes) and high pathogenic virus variants. In monocytes infected with low (PR8) and high pathogenic influenza viruses (FPV (H7N7) and H5N1), activity of the JAK-STAT1/2, JAK-STAT3 and TGFbeta pathways was distinctly lower when cells were infected by the virus, suggesting that partial suppression of the innate immune response may be part of the pathogenic mechanism used by the high pathogenic virus variants (FPV (H7N7), to evade the action of the immune response (FIG. 17).

Cellular Immune Response can be Measured on Different Immune Cell Types (FIGS. 18-20)

[0227] The method can be used to measure the cellular immune response on a variety of relevant immune cells or mixtures of immune cells as are frequently derived from blood samples, such as PBMCs. Cellular immune response was measured in whole blood of rotavirus infected patients, and activity of the JAK-STAT1/2 pathway was indicative of a cellular immune response (FIGS. 18 and 19). In both influenza patients and rotavirus patients the JAK-STAT1-2 INF 1 and INFII was higher during acute infection. JAK-STAT3 was also higher in the acute influenza infection but not the rotavirus.

[0228] CD8+ cells were used and informative in case of Dengue virus (FIG. 20), as well. CD4+ T cells and monocytes (CD16− inflammatory and CD16+ non-inflammatory) were used and informative in Yellow fever (see FIGS. 15 and 16). No high JAK-STAT1-2 pathway activity was measured, which is in agreement with the literature were was demonstrated that CD8 T cells from dengue patients become cytokine unresponsive due to TCR signaling insufficiencies. The measured JAK-STAT1/2 pathway activity was interferon type II dominant over type I.

Tuberculosis Vaccine (FIGS. 22 and 23)

[0229] Subjects were injected with M72/AS01E tuberculosis vaccine candidate at D0 and D30, whole blood samples were collected at D0, D30, D31, D37, D40, D44 and D47 and PBMC samples were collected at DO, D31, D44.

[0230] The tuberculosis vaccine consists of the adjuvant AS01E, containing MPL-A and immunostimulatory saponin QS-21, which induces an immune response characterized by the activation of interferon-gamma producing CD4+ T cells, and the production of antibodies.

[0231] The tuberculosis vaccination induces JAK-STAT1-2 type I and II IFN pathway activity after day 1 after vaccination in both blood sample types, which is in agreement with the authors of the paper, also the JAK-STAT3 and NFKB pathway activity was increased after day one of vaccination (see FIGS. 22 and 23).

Malaria Vaccination and Challenge—(FIG. 24)

[0232] Healthy malaria-naïve volunteers, randomized to two study arms participated in this study testing the efficacy of RTS, S and AdVac® malaria vaccine candidates. Study arm 1 (hereafter referred to as ARR), comprised of volunteers who received the AdVac® vaccine composed of Ad35 vector expressing full length CSP, as a primary immunization, followed by two doses of RTS, S/AS01E vaccine. The subjects in the second arm, received three doses of RTS, S/AS01 (RRR regimen). RTS, S/AS01E is a recombinant protein-based malaria vaccine.

[0233] The RTS, S vaccine was engineered using genes from the repeat and T-cell epitope in the pre-erythrocytic circumsporozoite protein (CSP) of the Plasmodium falciparum malaria parasite and a viral envelope protein of the hepatitis B virus (HBsAg), to which was added a chemical adjuvant (AS01E) to increase the immune system response. Participants in both study arms were vaccinated at 28-day intervals, and subjected to controlled malaria infection 21 days following the final immunization.

[0234] Both the RRR and ARR vaccination groups showed induced JAK-STAT1-2 type I and II IFN pathway activity after one day of vaccination, also the JAK-STAT3 pathway activity was increased after vaccinations (see FIG. 24). The ARR vaccine also resulted in significant increase in STAT1-2 after the two follow-up doses of RTS, S/AS01E.

[0235] The tuberculosis vaccine described above (FIGS. 22 and 23) is a bacterial vaccine with an adjuvant (AS01E; GSK). In general, no upregulation of the JAK-STAT1/2 cellular signaling pathway activity is observed from bacterial infections, as for example demonstrated by data obtained from sepsis patients. The upregulation of the JAK-STAT1/2 cellular signaling pathway activity observed in the dataset for the tuberculosis vaccine can be attributed to the adjuvant. The same adjuvant is used in the malaria vaccine described above (data relating to FIG. 24).

[0236] Therefore, it is envisioned that the immune response induced by a vaccine or vaccine efficacy can be determined based on the JAK-STAT1/2 cellular signaling pathway activity, if the vaccine comprises an adjuvant, in particular if the adjuvant is AS01, more in particular AS01E. It is envisioned that this response is independent of the vaccine type and thus the immune response induced by a vaccine or vaccine efficacy can be determined based on the JAK-STAT1/2 cellular signaling pathway activity for viral vaccines, bacterial vaccines and live attenuated vaccines.

Identification New JAK-STAT1/2 Target Genes.

[0237] Additional target genes were identified for the JAK-STAT1/2 cellular signaling pathway based on literature review and analysis on ground state truth datasets. Potential target genes were validated. The following genes were identified as JAK-STAT1/2 target genes which have been validated using the cellular signaling Pathway activity models described herein:

[0238] GBP1, TAP1, ISG15, APOL1, IFI6, IFIRM1, CXCL9, APOL2, IFIT2 and LY6E

Validation of 23, 16 and 3 Gene Models (FIGS. 25-31).

[0239] In order to validate the set of genes including the newly identified target genes for the JAK-STAT1/2 cellular signaling pathway, a selection of the datasets were analyzed again with the new 16 gene model, and put side by side with the old 23 gene model. It was found that the new 16 gene model performed at least equally. The data is indicated with 23g and 16g respectively.

[0240] To further support that also less genes can be used in the pathway models, random selection random selections of three genes were made as follows:

L1G3: USP18, APOL1, OAS1

L2G3: IRF1, TAP1, ISG15

L3G3: STAT1, IRF9, IFI6

L4G3: IRF7, GBP1, IFITM1

[0241] The data for datasets GSE22589 (FIG. 25), GSE29614 (FIG. 26), GSE34205 (FIG. 27), GSE35283 (FIG. 28), GSE43777 (FIG. 29), GSE50628 (FIG. 30) and GSE51997 (FIG. 31) was analyzed and compared. Similar trends were observed when using the randomly selected 3 gene models, although as expected the separation is less pronounced as compared to the 23 or 16 gene models.

Live Attenuated Yellow Fever Vaccine (FIG. 32)

[0242] The immune response can be followed over time. Healthy individuals were adminstered Yellow Fever vaccine (YFV-17, live attenuated yellow fever virus strain without adjuvant). In PBMCs the increase in JAK-STAT1/2 activity is seen to increase at day 3 and peak at day 7, to return to baseline levels at day 21 after administration of the vaccine.

COVID-19 (FIGS. 33-35)

Results

[0243] All results are based on pathway analysis of RNAseq data derived from whole blood samples.

1. Measuring Severity of COVID-19 (FIG. 33)

Dataset GSE157103

[0244] Dataset GSE157103 contains data from samples of COVID-19 patients and non-COVID-19 patients. Whole blood was collected from patients and RNA was isolated using LeukoLOCK Total RNA Isolation System[1]. RNA seq was performed.
Samples were annotated with respect to diagnosis (COVID-19 or non COVID-19) and severity (moderate to severe respiratory issues), ICU admission and mechanical ventilated status, for details see Overmyer et al.
For data analysis RNA sequencing-adapted signal transduction pathway analysis models for the AR, STAT1-2 and AP1 pathways were used, results shown in FIG. 33.
For the AR pathway (FIG. 33A), higher severity of disease was associated with higher AR pathway activity. This increase in AR pathway activity score was observed for both non-COVID patients who had other respiratory diseases and for COVID-19 patients, in which the mechanically ventilated patient group at the ICU had the highest AR pathway activity. The lowest AR pathway activity was found in the least severe group of non COVID patients, no ICU, and no mechanical ventilation.
For STAT1-2 pathway activity (FIG. 33C), lower pathway activities were associated with more severe disease, that is, lowest in ventilated patients at the ICU, intermediate in non-ventilated patients at the ICU, and highest in ventilated and non-ventilated patients at the general ward (not needing admission to the ICU). The lowest JAK-STAT1-2 pathway activity was found in both COVID and non-COVID patients who were mechanically ventilated at the ICU. Low JAK-STAT1/2 pathway activity in a virally infected patient indicates failure of the adaptive immune response to cope with the viral infection (Bouwman et al, Frontiers) [13]. It is known that COVID-19 patients may have impaired JAK-STAT1-2 pathway activity, and that very severe cases have acute respiratory distress syndrome (ARDS), necessitating admission at the ICU [14, p. 19]. From our pathway analysis it is clear that the patients without an impaired JAK-STAT1-2 pathway, resulting in high JAK-STAT1-2 pathway activity during COVID-19, could be treated outside the ICU with or without ventilation.
MAPK-AP1 pathway activity (FIG. 33B), showed no large differences between patients groups, but showed large spread in pathway activity scores within groups. This large spread for this signaling pathway could be due to the variety of comorbidities in patients with COVID-19, or different treatments that were administered and may have specifically affected activity of this pathway.
Conclusion from Analysis of Dataset GSE157103:
COVID-19 patients admitted to the ICU have more severe disease than those that can remain in a general ward; also at the ICU the patients who need ventilation generally have more severe disease than the patients who do not receive artificial ventilation. The results show that the level of JAK-STAT1/2 pathway activity can distinguish between patients on a general ward and patients at the ICU, and between ICU patients on and off artificial ventilation. The JAK-STAT1/2 pathway is a crucial pathway of which activity is needed to mount an adaptive immune response to a viral infection. The lower the activity (pathway activity score) of this pathway, the higher the severity of the disease. Therefore, these results show that the JAK-STAT1/2 pathway assay can be used to measure severity of COVID-19 in whole blood samples.
Based on earlier analysis results in patients with other viral infections, it is expected that measurement results of the activity of the JAK-STAT3 pathway will indicate the severity of the inflammatory condition in the acutely ill COVID-19 patients, mediated predominantly by the innate immune response, and therefore is expected to be inversely related to the activity of the JAK-STAT1/2 pathway.

Dataset GSE161777 (FIG. 34)

[0245] Dataset GSE161777 contains longitudinal data from 13 patients which were sampled at days 0, 2, 7, 10, 13 and/or at discharge. Whole blood samples were obtained, RNA isolated and RNA seq performed. One patient with mild disease was enrolled after recovery as a recovery control. 14 Healthy donors sampled at a single time point were included as controls. RNA was extracted from peripheral blood sampled at up to 5 time points per patient. At each sample point, a patient's disease trajectory, termed “pseudotime”, was categorized according to clinical parameters. To describe the heterogenous disease trajectories over time, a modified WHO ordinal scale (WHO, 2020) was used, which also takes into account several inflammatory markers (serum c-reactive protein [CRP], serum IL-6, and ferritin)[15]. The score (see FIG. 34A) was used to classify patients along their disease course and is proposed for use to monitor the disease (FIG. 34B). Pathway scores for all patients can be found in Table 2.
Comparing the various disease severity categories as defined in pseudotime (1-7 in FIG. 34) with respect to pathway activity, it is clear that (1) in the critically ill patients JAK-STAT 1/2 pathway activity is low (below 50, in the range of healthy controls) compared to pathway activity in patients with complicated or moderately severe disease (pathway activity score above 60, many around 80), and (2) during convalescence the corona virus-induced increased JAK-STAT1/2 pathway activity (above 60) (indicating an appropriate adaptive immune response to the virus) returns to the normal healthy level (below 50).
Conclusion from Dataset GSE161777
These results confirm that measurement of JAK-STAT1/2 pathway activity can be used to assess disease severity, where in the acute phase of the disease low JAK-STAT1/2 pathway activity means very severe disease with failure to mount a good adaptive immune response, while high JAK-STAT1/2 pathway activity indicates a good adaptive immune response. If the JAK-STAT 1/2 pathway activity was initially high and starts to decrease, this indicates that the patient is entering a reconvalescence phase.

2. Predicting Prognosis in COVID-19 Patients Using JAK-STAT1/2 Pathway Activity Dataset GSE161777 (FIG. 35).

[0246] As can be observed on the individual disease trajectories over time (pseudotime categories on the X-axis) (FIG. 35), repeated JAK-STAT1/2 pathway activity measurement can be used to predict prognosis in a patient with COVID 19.
When measuring JAK-STAT 1/2 pathway activity at least twice sequentially in a patient with covid-19, addition of the pathway activity score to conventional clinical parameters of the patient (here defined as pseudotime parameters 1-7) is expected to improve prediction of prognosis of the patient. For example, if in an clinically acute disease phase JAK-STAT1/2 pathway activity decreases over a few days, this is likely to predict a bad prognosis; in contrast, if the JAK-STAT1/2 pathway activity shows an increase over that time period, this indicates a good adaptive immune response, with expected good prognosis. If two JAK-STAT1/2 measurements are performed while the patient is showing improvement according to conventional clinical parameters, an increase in pathway activity score indicates that the patient is still in the acute phase but mounts an adequate adaptive immune response with a good prognosis; while a decrease in pathway activity score indicates that that patient enters the reconvalescence phase with a good prognosis.
With respect to activity scores for AR and MAPK pathways, no difference was observed between critically ill and complicated/moderately ill patients, however pathway activities were increased compared to healthy controls, and decreased during reconvalescence to normal values (FIG. 35). Activity of these pathways are not associated with activity of the adaptive immune system, but indicate activity of the inflammatory innate immune response. Measurements of activity cannot be used to assess severity during acute COVID-19 disease, but can be used for monitoring changes in disease severity, and based on that, predicting prognosis of the patient (see below).

[0247] 1. Detailed Description of the Dataset Analysis Results, Per Pseudotime Severity Category

[0248] a. Critically Ill Patients

To determine if the severity of disease (pseudo-time severity categories 1-7 in FIG. 35) could be linked to pathway scores we analyzed the RNA sequencing data. In FIG. 35A and Table 2, the AP1, AR and STAT1-2 pathway activity scores of critically ill patients are shown, healthy individuals are plotted as HC.
All patients in this group showed low STAT1-2 pathway activity, further confirming that low JAK-STAT1/2 pathway activity is associated with severe disease
In non survivor patient 2, AR and AP1 pathway activity increased over time (pseudotime category 1 and 2).
Non-survivor Patient 3 had low pathway activity scores for all three pathways, probably associated with an impaired immune system due to chemotherapy treatment.
Survivor-patient 12 showed decreasing pathway activity scores for the AP1 and AR pathways over time which was not seen for non-survivor patient 2 and 3. The decrease in AR and MAPK-AP1 pathway activity was associated with recovery. Lower activity of these pathways are likely to be associated with less severe inflammatory disease and activity of the innate immune system.

[0249] b. Complicated Patients

Complicated patients showed higher AP1, AR and STAT1-2 pathway activity compared to healthy controls (HC), and pathway activities decreased over the pseudotime during reconvalescence phase. The lowest pathway activity scores were observed in the recovery phase, see FIG. 35B, Table 2.

[0250] c. Moderately Ill Patients

In the moderate ill patient group higher AR, MAPK-AP1 and JAK-STAT1/2 pathway activity scores compared to healthy controls HC) were only found with the pseudo score 4 and not in later pseudo scores (5-7) in which the patients are recovering and the pathway activities are normalized towards the healthy control levels, FIGS. 35C and A,

[0251] d. Late Convalescence/Recovery Patients

In the Late convalescence/recovery patient group (FIG. 35D, Table 2), patient #6 showed decreasing pathway activity scores for the MAPK-AP1, AR and JAK-STAT1-2 pathways over the pseudotime.
For patient #11 several samples were taken during the same pseudotime phase 6. The MAPK-AP1 and AR pathway activity scores decreased over pseudotime, while JAK-STAT1-2 pathway activity scores showed a large variation during pseudotime 6 and entered the normal range during recovery (pseudotime 7).

Changes in AR and MAPK Pathway Activity During Disease Progression and Recovery (FIG. 36).

[0252] The incremental phase in AR and MAPK-AP1 pathway activity was observed in patient 4 and 10 who showed increased AP1 and AR pathway activity (and persistent high JAK-STAT1-2 pathway activity) in pseudotime nr 1 to 3. Such an increase was also seen in patient 1 who was re-admitted and in patient 2 (non-survivor). decreased pathway activities of patients 1, 4 and 10 was already seen at pseudotime 3 for patient 4 and pseudotime 5 for patients 1 and 10. See FIG. 36 and FIG. 34B for patient groups.

Dataset GSE161731

[0253] Samples from subjects with COVID-19 were assigned to three groups based on time from symptom onset (early ≤10 days, middle 11-21 days, late >21 days). For comparison, we profiled banked blood samples from patients presenting to the emergency department with acute respiratory infection (ARI) due to seasonal coronavirus (n=49), influenza (n=17) or bacterial pneumonia (n=23), and matched healthy controls (n=19).
Pathway activities were determined for each sample, per group average pathway activities and mean values are indicated in Table 3. This dataset further confirms that JAK-STAT1/2 pathway activity is initially high and then decreases in COVID-19 infections, likely due to the initial adaptive immune response which then subsides. The dataset further also confirms that JAK-STAT1/2 pathway activity is generally high in viral infections while not elevated in bacterial infections. Interestingly the JAK-STAT1/2 activity is much less elevated in COVID-19 patients compared to patients with seasonal coronavirus infections or influenza, suggesting a adaptive immune suppressive effect from SARS-CoV2. This could explain the generally more severe symptoms and further emphasize that those patients with lower JAK-STAT1/2 pathway activity (suggesting lower/no adaptive immune response) tend to have much more severe symptoms.

[0254] The dataset further shows a strong increase in AR pathway activity specifically in bacterial infected patients. Therefore the AR pathway activity can further be used to distinguish between bacterial and viral infected patients based on a blood sample.

REFERENCES

[0255] 1. Wang D, Hu B, Hu C, Zhu F, Liu X, Zhang J, et al. Clinical Characteristics of 138 Hospitalized Patients With 2019 Novel Coronavirus-Infected Pneumonia in Wuhan, China. JAMA. 2020 Feb. 7; [0256] 2. Zhou F, Yu T, Du R, Fan G, Liu Y, Liu Z, et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. Lancet. 2020 28;395(10229):1054-62. [0257] 3. Verhaegh W, van Ooijen H, Inda M A, Hatzis P, Versteeg R, Smid M, et al. Selection of personalized patient therapy through the use of knowledge-based computational models that identify tumor-driving signal transduction pathways. Cancer Res. 2014 Jun. 1;74(11):2936-45. [0258] 4. van Ooijen H, Hornsveld M, Dam-de Veen C, Velter R, Dou M, Verhaegh W, et al. Assessment of Functional Phosphatidylinositol 3-Kinase Pathway Activity in Cancer Tissue Using Forkhead Box-O Target Gene Expression in a Knowledge-Based Computational Model. Am J Pathol. 2018 September;188(9):1956-72. [0259] 5. Stolpe A van de, Holtzer L, van Ooijen H, de Inda M A, Verhaegh W. Enabling precision medicine by unravelling disease pathophysiology: quantifying signal transduction pathway activity across cell and tissue types. Sci Rep. 2019 Feb. 7;9(1):1603. [0260] 6. Verhaegh W, Van de Stolpe A. Knowledge-based computational models. Oncotarget. 2014 Jul. 30;5(14):5196-7. [0261] 7. van de Stolpe A. Quantitative Measurement of Functional Activity of the PI3K Signaling Pathway in Cancer. Cancers. 2019 March;11(3):293. [0262] 8. van Boxtel W, Verhaegh G W, van Engen-van Grunsven I A, van Strijp D, Kroeze L I, Ligtenberg M J, et al. Prediction of clinical benefit from androgen deprivation therapy in salivary duct carcinoma patients. Int J Cancer. 2019 Nov. 19; [0263] 9. Inda M A, Blok E J, Kuppen P J K, Charehbili A, den Biezen-Timmermans E C, van Brussel A, et al. Estrogen Receptor Pathway Activity Score to Predict Clinical Response or Resistance to Neoadjuvant Endocrine Therapy in Primary Breast Cancer. Mol Cancer Ther. 2020 February;19(2):680-9. [0264] 10. Sieuwerts A M, Inda M A, Smid M, van Ooijen H, van de Stolpe A, Martens J W M, et al. ER and PI3K Pathway Activity in Primary ER Positive Breast Cancer Is Associated with Progression-Free Survival of Metastatic Patients under First-Line Tamoxifen. Cancers (Basel). 2020 Mar. 27;12(4). [0265] 11. Stolpe A V D, Verhaegh W, Doom A van, Noel G, Gu-Trantien C, Willard-Gallo K. Abstract 2371: Breast cancer induces tolerogenic state of healthy activated CD4+ lymphocytes, characterized by reduced PI3K, NFKB, JAK-STAT, Notch, and increased TGFβpathway activity. Cancer Res. 2019 Jul. 1;79 (13 Supplement):2371-2371. [0266] 12. K. A. Overmyer et al., “Large-Scale Multi-omic Analysis of COVID-19 Severity,” cels, vol. 12, no. 1, pp. 23-40.e7, January 2021, doi: 10.1016/j.cels.2020.10.003. [0267] 13. W. Bouwman, W. Verhaegh, L. Holtzer, and A. van de Stolpe, “Measurement of Cellular Immune Response to Viral Infection and Vaccination,” Front Immunol, vol. 11, p. 575074, 2020, doi: 10.3389/fimmu.2020.575074. [0268] 14. T. Matsuyama, S. P. Kubli, S. K. Yoshinaga, K. Pfeffer, and T. W. Mak, “An aberrant STAT pathway is central to COVID-19,” Cell Death & Differentiation, vol. 27, no. 12, Art. no. 12, December 2020, doi: 10.1038/s41418-020-00633-7. [0269] 15. “Longitudinal Multi-omics Analyses Identify Responses of Megakaryocytes, Erythroid Cells, and Plasmablasts as Hallmarks of Severe COVID-19: Immunity.” https://www.cell.com/immunity/fulltext/S1074-7613(20)30504-5?_returnURL=https %3A %2F %2Flinkinghub.elsevier.com %2Fretrieve %2Fpii %2FS107476 1320305045%3Fshowall %3Dtrue (accessed Mar. 29, 2021). [0270] 16. M. K. Angele, S. Pratschke, W. J. Hubbard, and I. H. Chaudry, “Gender differences in sepsis,” Virulence, vol. 5, no. 1, pp. 12-19, January 2014, doi: 10.4161/viru.26982. [0271] 17. M. R. Gubbels Bupp and T. N. Jorgensen, “Androgen-Induced Immunosuppression,” Front Immunol, vol. 9, p. 794, 2018, doi: 10.3389/fimmu.2018.00794.

TABLE-US-00001 TABLE 1 Patient material Virus/ Sample of infected bacteria Dataset type cells type Sample information Reference GSE84331 CD8 T Patient Dengue PBMCs isolated from https://www.ncbi.nlm.nih.gov/ Cell hemorrhagic peripheral whole blood of pubmed/27707928 fever dengue patients and healthy controls were stained with relevant antibodies at 4° C. for 30 mins, washed thoroughly and suspended in PBS containg 2% FCS and immediately sorted on a BD FACS Aria III (Becton and Dickenson) with high forward scatter gates to account for the larger blasting effector lymphocytes. CD3+ CD8+ CD45RA+ CCR7+ naïve CD8 T cells and CD3+ CD8+ HLA-DR+ CD38+ activated CD8 T cells were isolated up to a purity of 99% CD8 T Patient Dengue Cell fever GSE119322 Dendritic Patient Hepatitis Mononuclear cells were https://www.ncbi.nlm.nih.gov/ cells B isolated from side-scatter and pubmed/24616721 forward-scatter gating using fluorescence-activated cell sorting (FACS), the lineage- negative HLA-DR-positive fraction was extracted, and was further divided into CD123-positive plasmacytoid DCs (pDCs) and CD11c-positive myeloid DCs (mDCs). Cell sorting was performed, and the number of DCs was measured and surface markers were analyzed. RNA was extracted from sorted pDCs and subjected to gene expression analysis using Affymetrix Human 133U Plus 2.0 gene chip. GSE52081 Dendritic infected Newcastle Human peripheral blood https://www.ncbi.nlm.nih.gov/ cells cells disease mononuclear cells were pubmed/24616721 virus isolated from buffy coats by Ficoll density gradient centrifugation (Histopaque, Sigma Aldrich) at 1450 r.p.m. and CD14+ monocytes were immunomagnetically purified by using a MACS CD14 isolation kit (Miltenyi Biotech). Monocytes were then differentiated into naïve DCs by 5-6 day incubation at 37° C. and 5% CO2 in DC growth media, which contains RPMI Medium 1640 (Invitrogen/Gibco) supplemented with 10% fetal calf serum (Hyclone), 2 mM of l-glutamine, 100 U/ml penicillin and 100 g/ml streptomycin (Pen/Strep) (Invitrogen), 500 U/ml hGM- CSF (Preprotech) and 1000 U/ml hIL-4 (Preprotech). Antiviral activated dendritic cells (AVDCs) were generated by employing a trans-well system. The trans-well system consists of an upper and a lower chamber separated by a 0.4 μm PET membrane (Millipore) that allows diffusion of cytokines and chemokines through the membrane but avoids the interaction of the cells in both chambers. To generate the AVDCs, naïve DCs were infected as described above. After the 40 minutes incubation, the cells were washed with PBS, and cultured in the trans-well system. Infected and non- infected DCs were allocated in the upper and lower chamber respectively. GSE35283 Monocytes infected influenza Human monocytes were https://www.ncbi.nlm.nih.gov/ cells low PR8 isolated from buffy coats of pubmed/23445660 unrelated healthy blood donors. Cells were cultivated in Teflon bags in McCoy's modified medium (Biochrom AG, Berlin, Germany) supplemented with 1% glutamine, 1% penicillin- streptomycin and 15% fetal bovine serum overnight. Monocytes were infected with low (PR8) and high pathogenic influenza viruses (FPV (H7N7) and H5N1) Monocytes infected influenza cells high FPV H7N7 Monocytes infected influenza cells high H5N1 GSE34205 PBMCs Patient RSV acute RSV or Influenza https://www.ncbi.nlm.nih.gov/ infection, children of median pubmed/22398282 age 2.4 months (range 1.5-8.6) hospitalized with acute RSV and influenza virus infection were offered study enrollment after microbiologic confirmation of the diagnosis. Blood samples were collected from them within 42-72 hours of hospitalization. PBMCs were isolated within 6 h of sample collection by density gradient centrifugation using ficoll-hypaque and lysed in RLT reagent (Qiagen) with β- mercaptoethanol. Samples were run blind and in batches by the same laboratory team to ensure standardization of quality and handling. From 2-5 ug of total RNA, cDNA was generated as a template for single-round in vitro transcription with biotin- labeled nucleotides using the Affymetrix cDNA Synthesis and In Vitro Transcription kits (Affymetrix Inc.). PBMCs Patient Influenza GSE43777 PBMCs Patient Denque DF (dengue fever) and DHF https://www.ncbi.nlm.nih.gov/ (dengue hemorrhagic fever) pubmed/23875036 cases were used to study gene expression during the course of dengue acute illnessOnly if DENV infection was confirmed by RT-PCR, then serial blood samples were collected at 24, 48 and 72 hours following the initial sample, and one to two samples within 0-72 hours post-fever defervescence and one sample at ≥20 days (convalescent period) following the initial sample. Separation of plasma and PBMCs was performed by gradient centrifugation over Histopaque-Ficoll (Sigma, St. Louis, MO). GSE50628 PBMCs Patient Influenza The gene expression profiles https://www.ncbi.nlm.nih.gov/ H1N1 in the peripheral whole blood pubmed/24464411 with influenza A(H1N1)pdm09 or rotavirus gastroenteritis were examined. Whole blood samples were collected from patients in the acute phase of the disease and in the recovery phase. For patients with complex seizures, the blood samples were collected on the day of admission (acute phase: Flu, 1-3 days from disease onset; Rota, 2-4 days from disease onset) and on the day of discharge (recovery phase: Flu, 4-9 days from disease onset; Rota, 7-11 days from disease onset). After sample collection, the PAXgene tubes were incubated at room temperature for 2 h and then stored at −80° C. until RNA extraction. Total RNA was isolated using the PAXgene Blood RNA Kit (Qiagen). PBMCs Patient rota virus GSE69606 PBMCs Patient RSV patients with acute RSV https://www.ncbi.nlm.nih.gov/ infections, divided into mild, pubmed/26162090 moderate and severe disease. From moderate and severe diseased patients recovery samples were obtained as well. Peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation (Lymphoprep, Axis Shield, Norway), counted and subsequently stored in Trizol reagent (Invitrogen, The Netherlands) at −80° C. in the same laboratory by the same team for both cohorts. RNA from PBMC was extracted using Trizol (Invitrogen Life Technologies) according to the manufacturer's protocol. Total RNA was isolated using the RNeasy Minikit (Qiagen). GSE6269 PBMCs Patient Influenza Peripheral blood samples from https://www.ncbi.nlm.nih.gov/ pediatric patients with acute pubmed/17105821 infections of Influenza, S. pneumoniae or S. aureus. All blood samples were collected in acid-citrate-dextrose tubes (BD Vacutainer, Becton Dickinson, Franklin Lakes, NJ) at the CMC and immediately delivered at room temperature to the Baylor Institute for Immunology Research (Dallas, TX) for processing. Peripheral blood mononuclear cells (PBMCs) from 3 to 4 mL blood were isolated via Ficoll gradient and immediately lysed in RLT reagent (Qiagen, Valencia, CA) with β-mercaptoethanol (BME) and stored at −80° C. Total RNA was isolated using the RNeasy kit (Qiagen) according to the manufacturer's instructions, and RNA integrity was assessed by using an Agilent 2100 Bioanalyzer (Agilent, Palo Alto, CA). PBMCs Patient S. pneumoniae PBMCs Patient S. aureus GSE102459 Whole Patient tuberculosis All eligible participants https://www.ncbi.nlm.nih.gov/ blood were stipulated to receive pmc/articles/PMC5879450/ and the candidate vaccine BPMCs M72/AS01.sub.E (referred to as M72/AS01 in the article; GSK, Rixensart, Belgium) by intramuscular injection at Days 0 and 30. PBMC were collected at Days 0, 31, and 44; and with WB-derived samples on Days 0, 30, 31, 37, 40, 44, and 47. BMCs were separated on Lymphoprep gradients, washed, counted by flow cytometry, frozen and further stored in liquid nitrogen until time of further evaluation. At least 10 ml of blood for WB gene expression analysis was collected in PAXgene tubes. GSE89292 PBMCs Patient Malaria Study arm 1 (ARR), https://www.ncbi.nlm.nih.gov/ comprised of volunteers pmc/articles/PMC5338562/ who received the AdVac vaccine composed of Ad35 vector expressing full-length CSP, as a primary immunization, was followed by two doses of RTS, S/AS01 vaccine. The subjects in the second arm, , received three doses of RTS, S/AS01 (RRR regimen). Participants in both study arms were vaccinated at 28-d intervals, and subjected to CHMI 21 d following the final immunization. mRNA was isolated from frozen PBMCs GSE4607 PBMCs Patient bacterial Children <10 years of age https://www.ncbi.nlm.nih.gov/ infections admitted to the pediatric pubmed/18511707 intensive care unit and meeting the criteria for either SIRS or septic shock were eligible for the study. SIRS and septic shock were defined based on pediatric-specific criteria. We did not use separate categories of “sepsis” or “severe sepsis”. Patients meeting criteria for “sepsis” or “severe sepsis” were placed in the categories of SIRS and septic shock, respectively, for study purposes. Total RNA was extracted from whole blood samples using the PaxGene Blood RNA System (PreAnalytiX, Qiagen/Becton Dickson) according the manufacturer's specifications. Quality control steps: RNA quality was assessed by using the Agilent bioanalyzer (Agilent Technologies, Palo Alto, CA) and only those samples with 28S/18S ratios between 1.3 and 2 were subsequently used GSE13486 PBMCs Patient YFV-17, vaccination in healthy https://pubmed.ncbi.nlm.nih.gov/ live individuals with Yellow Fever 19029902/ attenuated vaccine yellow fever virus strain without adjuvant GSE157103 Whole COVID-19 COVID-19 Large-scale Multi-omic https://pubmed.ncbi.nlm.nih.gov/ blood patients Analysis of COVID-19 Severity 33096026/ and non- COVID-19 patients GSE161777 peripheral 13 patients COVID-19 Longitudinal multi-omics https://pubmed.ncbi.nlm.nih.gov/ blood were identifies responses of 33296687/ sampled megakaryocytes, erythroid at days 0, cells and plasmablasts as 2, 7, 10, hallmarks of severe COVID-19 13 and/or trajectories [sequencing] at discharge. GSE161731 Whole Expression COVID-19, Dysregulated transcriptional https://pubmed.ncbi.nlm.nih.gov/ blood profiling seasonal responses to SARS-CoV-2 in 33597532/ by high coronavirus, the periphery support novel throughput influenza, diagnostic approaches sequencing bacterial pneumonia

TABLE-US-00002 TABLE 2 pseudo- title time remission AP1 AR STAT1 Healthy control 01 0 Healthy 10.8 6.3 21.9 Healthy control 02 0 Healthy 13.5 9.5 36.9 Healthy control 03 0 Healthy 15.2 9.0 40.1 Healthy control 04 0 Healthy 14.3 9.7 34.0 Healthy control 05 0 Healthy 7.8 4.6 21.5 Healthy control 06 0 Healthy 18.6 11.1 77.0 Healthy control 07 0 Healthy 13.7 5.5 69.6 Healthy control 08 0 Healthy 11.2 1.5 44.9 Healthy control 09 0 Healthy 13.7 8.5 34.2 Healthy control 10 0 Healthy 13.3 8.1 24.0 Healthy control 11 0 Healthy 9.4 4.7 25.0 Healthy control 11 0 Healthy 9.6 5.1 25.0 Healthy control 12 0 Healthy 13.5 8.5 20.4 Healthy control 12 0 Healthy 13.4 8.8 20.4 Healthy control 13 0 Healthy 12.1 6.6 23.0 Healthy control 13 0 Healthy 12.0 6.7 23.1 Healthy control 14 0 Healthy 10.7 6.7 16.8 Healthy control 14 0 Healthy 10.8 6.7 16.9 patient1: COVID19 4 Remission 24.5 9.8 83.2 (Remission) t = 4 patient1: COVID19 4 Remission 23.0 9.9 81.2 (Remission) t = 4 patient1: COVID19 4 Remission 30.5 17.4 70.9 (Remission) t = 4 patient1: COVID19 4 Remission 27.3 17.4 70.5 (Remission) t = 4 patient1: COVID19 5 Remission 15.8 11.3 34.5 (Remission) t = 5 patient1: COVID19 5 Remission 15.9 10.8 35.0 (Remission) t = 5 patient1: COVID19 6 Remission 12.7 6.0 27.2 (Remission) t = 6 patient1: COVID19 6 Remission 12.7 6.7 27.4 (Remission) t = 6 patient1: COVID19 6 Remission 12.1 6.1 27.2 (Remission) t = 6 patient1: COVID19 6 Remission 12.2 6.2 26.7 (Remission) t = 6 patient10: COVID19 1 Remission 24.1 16.6 66.1 (Remission) t = 1 patient10: COVID19 1 Remission 26.9 15.6 70.7 (Remission) t = 1 patient10: COVID19 3 Remission 24.7 19.9 80.8 (Remission) t = 3 patient10: COVID19 3 Remission 24.7 19.6 80.9 (Remission) t = 3 patient10: COVID19 3 Remission 15.4 9.1 66.4 (Remission) t = 3 patient10: COVID19 3 Remission 15.4 9.5 67.2 (Remission) t = 3 patient10: COVID19 4 Remission 16.6 15.1 62.7 (Remission) t = 4 patient10: COVID19 4 Remission 17.0 15.2 64.3 (Remission) t = 4 patient10: COVID19 5 Remission 10.4 2.6 56.6 (Remission) t = 5 patient10: COVID19 5 Remission 10.2 2.2 56.6 (Remission) t = 5 patient11: COVID19 6 Remission 17.4 14.9 27.1 (Remission) t = 6 patient11: COVID19 6 Remission 17.6 14.9 26.7 (Remission) t = 6 patient11: COVID19 6 Remission 17.6 16.1 34.2 (Remission) t = 6 patient11: COVID19 6 Remission 17.4 15.8 33.6 (Remission) t = 6 patient11: COVID19 6 Remission 17.6 16.2 44.2 (Remission) t = 6 patient11: COVID19 6 Remission 17.5 15.9 45.3 (Remission) t = 6 patient11: COVID19 6 Remission 18.4 15.1 53.9 (Remission) t = 6 patient11: COVID19 6 Remission 18.2 14.8 53.6 (Remission) t = 6 patient12: COVID19 2 Remission 17.1 7.2 41.7 (Remission) t = 2 patient12: COVID19 2 Remission 16.8 7.3 41.9 (Remission) t = 2 patient12: COVID19 3 Remission 2.0 0.0 13.5 (Remission) t = 3 patient12: COVID19 3 Remission 2.2 0.0 13.6 (Remission) t = 3 patient12: COVID19 5 Remission 3.8 0.3 17.9 (Remission) t = 5 patient12: COVID19 5 Remission 4.4 0.2 17.8 (Remission) t = 5 patient12: COVID19 5 Remission 0.6 0.1 8.7 (Remission) t = 5 patient12: COVID19 5 Remission 0.7 0.1 8.4 (Remission) t = 5 patient12: COVID19 5 Remission 1.5 0.2 14.5 (Remission) t = 5 patient12: COVID19 5 Remission 1.8 0.3 14.3 (Remission) t = 5 patient13: COVID19 4 Remission 25.5 14.8 70.2 (Remission) t = 4 patient13: COVID19 4 Remission 24.4 14.4 67.1 (Remission) t = 4 patient13: COVID19 5 Remission 11.6 5.5 32.5 (Remission) t = 5 patient13: COVID19 5 Remission 11.0 5.7 31.9 (Remission) t = 5 patient14: COVID19 4 Remission 21.8 7.6 74.1 (Remission) t = 4 patient14: COVID19 4 Remission 21.7 7.6 75.3 (Remission) t = 4 patient14: COVID19 5 Remission 22.3 14.3 50.1 (Remission) t = 5 patient14: COVID19 5 Remission 20.7 14.0 51.0 (Remission) t = 5 patient2: COVID19 (No 1 No 18.1 9.2 29.3 Remission) t = 1 Remission patient2: COVID19 (No 1 No 18.3 9.2 29.3 Remission) t = 1 Remission patient2: COVID19 (No 2 No 21.7 13.5 29.7 Remission) t = 2 Remission patient2: COVID19 (No 2 No 21.5 13.5 29.6 Remission) t = 2 Remission patient3: COVID19 (No 2 No 13.5 6.1 24.3 Remission) t = 2 Remission patient3: COVID19 (No 2 No 13.5 5.8 24.2 Remission) t = 2 Remission patient4: COVID19 1 Remission 29.1 14.3 80.3 (Remission) t = 1 patient4: COVID19 1 Remission 28.1 14.1 80.1 (Remission) t = 1 patient4: COVID19 3 Remission 27.7 18.6 82.5 (Remission) t = 3 patient4: COVID19 3 Remission 28.4 18.1 80.0 (Remission) t = 3 patient4: COVID19 5 Remission 15.3 8.8 30.7 (Remission) t = 5 patient4: COVID19 5 Remission 15.0 8.4 31.5 (Remission) t = 5 patient5: COVID19 1 Remission 19.8 12.5 68.3 (Remission) t = 1 patient5: COVID19 1 Remission 19.8 12.5 68.1 (Remission) t = 1 patient5: COVID19 3 Remission 16.6 8.1 48.4 (Remission) t = 3 patient5: COVID19 3 Remission 16.8 7.8 48.6 (Remission) t = 3 patient5: COVID19 6 Remission 5.7 2.1 9.6 (Remission) t = 6 patient5: COVID19 6 Remission 6.3 1.5 9.4 (Remission) t = 6 patient5: COVID19 6 Remission 3.7 2.7 6.5 (Remission) t = 6 patient5: COVID19 6 Remission 3.6 2.5 6.3 (Remission) t = 6 patient6: COVID19 5 Remission 24.5 10.9 69.0 (Remission) t = 5 patient6: COVID19 5 Remission 25.1 10.8 68.5 (Remission) t = 5 patient6: COVID19 5 Remission 25.6 12.8 55.6 (Remission) t = 5 patient6: COVID19 5 Remission 23.9 12.6 55.2 (Remission) t = 5 patient6: COVID19 6 Remission 11.0 6.2 15.6 (Remission) t = 6 patient6: COVID19 6 Remission 11.0 5.7 15.7 (Remission) t = 6 patient6: COVID19 7 Remission 9.5 4.0 20.6 (Remission) t = 7 patient6: COVID19 7 Remission 9.5 3.1 20.4 (Remission) t = 7 patient7: COVID19 7 Remission 14.0 11.8 39.1 (Remission) t = 7 patient7: COVID19 7 Remission 13.8 11.7 40.4 (Remission) t = 7 patient8: COVID19 3 Remission 22.7 12.6 71.6 (Remission) t = 3 patient8: COVID19 3 Remission 21.8 12.7 76.9 (Remission) t = 3 patient9: COVID19 4 Remission 15.7 9.2 30.1 (Remission) t = 4 patient9: COVID19 4 Remission 15.8 9.1 31.2 (Remission) t = 4 patient9: COVID19 5 Remission 10.5 7.7 16.3 (Remission) t = 5 patient9: COVID19 5 Remission 10.5 7.9 16.5 (Remission) t = 5 patient9: COVID19 6 Remission 5.1 4.9 8.2 (Remission) t = 6 patient9: COVID19 6 Remission 5.2 4.9 8.1 (Remission) t = 6

TABLE-US-00003 TABLE 3 AP1 AR ER FOXO HH NOTCH STAT1-2 TGFbeta PI3K Average COVID-19 all early 25.4 11.0 5.3 33.9 6.8 10.4 76.4 12.6 66.1 middle 25.0 12.0 3.9 39.1 6.2 7.2 66.2 13.0 60.9 late 25.0 11.0 3.5 41.0 5.9 7.3 64.0 13.2 59.0 STDEV COVID-19 all early 1.4 3.0 2.0 7.0 2.9 4.0 8.8 1.7 7.0 middle 1.3 2.3 1.2 5.3 1.5 2.9 6.0 2.0 5.3 late 1.5 1.4 1.3 4.8 1.2 2.1 6.6 1.6 4.8 Average hospitalized early 25.3 10.3 6.0 29.7 6.1 9.0 77.8 12.1 70.3 middle 28.0 11.6 4.8 38.0 9.0 7.8 70.8 11.8 62.0 STDEV early 1.6 4.3 1.4 4.1 3.0 3.3 9.6 1.8 4.1 hospitalized middle 5.4 3.0 2.9 7.9 3.2 2.3 14.1 2.8 7.9 Average non early 25.4 11.3 5.0 35.9 7.1 11.1 75.8 12.9 64.1 hospitalized middle 25.0 12.0 3.9 39.1 6.2 7.2 66.2 13.0 60.9 late 25.0 11.0 3.5 41.0 5.9 7.3 64.0 13.2 59.0 STDEV early 1.3 2.3 2.2 7.4 2.9 4.2 8.7 1.7 7.4 middle 1.3 2.3 1.2 5.3 1.5 2.9 6.0 2.0 5.3 late 1.5 1.4 1.3 4.8 1.2 2.1 6.6 1.6 4.8 bacterial average 26.5 16.5 5.5 48.2 2.9 3.2 56.0 13.8 51.8 STDEV 3.5 3.2 1.8 5.0 1.8 2.5 11.8 2.7 5.0 Viral average 25.7 11.8 4.4 37.4 5.2 11.4 86.8 13.3 62.6 STDEV 1.6 2.3 1.3 6.2 2.3 2.0 5.0 1.0 6.2 Heathy average 24.5 11.3 2.8 40.2 4.9 6.9 69.8 13.4 59.8 STDEV 1.7 2.2 1.4 6.8 1.7 3.4 8.5 2.1 6.8