Plasma cell-free RNA signatures of inflammatory syndromes in children

Abstract

The current disclosure is directed to cell-free RNA (cfRNA) profiling to compare host immune and cellular injury responses associated with different inflammatory and / or infectious syndromes in children. The current disclosure describes methods of differentiating and detecting different inflammatory and / or infectious syndromes in children. In one aspect, the current disclosure is directed to a method comprising: a) obtaining cell-free RNA from a biological sample of a subject, wherein the biological sample is a plasma sample or a serum sample; b) determining a profile of a panel of biomarkers in the cell-free RNA; and c) determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers.

Description

PLASMA CELL-FREE RNA SIGNATURES OF INFLAMMATORYSYNDROMES IN CHILDRENCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Application Serial No. 63 / 560,355, filed March 1, 2024, the entire contents of which are incorporated herein by reference.STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0002] This disclosure was made with government support under a research project supported by Grant R33HD105618-03S1 awarded by the U.S. National Institute of Health. The government has certain rights in this invention.BACKGROUND OF THE INVENTION

[0003] The differential diagnosis of inflammatory syndromes in children is complex owing to their overlapping clinical manifestations, non-specific symptoms, and developmental age-related barriers to communication. These challenges often result in delayed or inaccurate diagnoses, thereby impeding effective clinical management and increasing the risk of long-term adverse health effects. A key example is Kawasaki disease (KD), an inflammatory syndrome of unknown etiology that primarily affects children under five years of age. KD is often misdiagnosed because significant clinical signs overlap between KD and other inflammatory and / or infectious conditions. Accurate diagnosis is critical, as KD patients who do not receive intravenous immunoglobulin (IVIG) early in the course of illness have a substantially increased risk of developing coronary artery aneurysms, making KD the leading cause of acquired heart disease in children. Thus, there is a clear need for accurate molecular tests for inflammatory conditions such as KD to inform timely and appropriate treatment.

[0004] Currently, the differential diagnosis of pediatric inflammatory syndromes relies on clinical assessment of signs and symptoms and results from a broad array of laboratory tests. Culture-based and molecular assays are routinely used to identify viral and bacterial pathogens,but these tests do not interrogate the host response and hence are unable to differentiate between infectious and non-infectious conditions. Serologic metabolic and antigen biomarkers used for diagnosis often lack specificity. To address these limitations, recent studies have explored the use of whole blood RNA transcriptome profiling of the human host response to assess disease severity and differentiate among inflammatory conditions including KD, MIS-C, viral, and bacterial infection. However, while the whole blood profile is indicative of the host immune response, it provides little information regarding the extent of inflammation-related cell injury or death in solid organ tissues. In contrast, cell-free nucleic acids in plasma, including cell-free DNA (cfDNA) and cell-free RNA (cfRNA) are promising analytes for evaluating inflammation as they are released by dead or dying cells originating from both the bloodstream and solid tissues. Many recent studies have explored the use of cell-free nucleic acids for monitoring of pregnancy, cancer, transplantation and infection, yet the potential of cell-free nucleic acids for the differential diagnosis of inflammatory syndromes remains largely unexplored.SUMMARY OF THE INVENTION

[0005] The current disclosure is directed to cell-free RNA (cfRNA) profiling to compare host immune and cellular injury responses associated with different inflammatory and / or infectious syndromes in children. The current disclosure describes methods of differentiating and detecting different inflammatory and / or infectious syndromes in children.

[0006] In one aspect, the current disclosure is directed to a method comprising: a) obtaining cell-free RNA from a biological sample of a subject, wherein the biological sample is a plasma sample or a serum sample; b) determining a profile of a panel of biomarkers in the cell-free RNA; and c) determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers.

[0007] In some embodiments, the step of determining a profile of a panel of biomarkers comprises performing nucleotide sequencing of the cell-free RNA. In some embodiments, the step of determining a profile of a panel of biomarkers comprises measuring the levels of RNAs corresponding to the biomarkers in the panel. In some embodiments, the panel of biomarkers comprises one, more, or all (including, e.g., at least 30, at least 50, at least 75, or at least 100) of genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12,PABPC4, ANXA11 , BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI , FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK, ANXA6, VCAN, PGD, PHACTR2, S100A8, PFN1, VIM, MIB1, TCF7, NRIP1, CSF3R, MX1, BIRC2, PTGES3, POU2F2, FKBP5, HDLBP, H2BC9, DARS1, AKNA, LCP1, SH3BP5, CCDC88C, DST, EIF4B, MVP, GRB2, HIP1, CD74, NFE2L1, WIPF1, PEAK1, CSK, TBC1D14, PTMA, KDM6B, LYZ, NUCKS1, RAPGEF1, MT-TL1, HIF1A, TXNIP, YBX1, GPI, PKP4, RPL21, RABEP1, TDP2, MAP4K4, GAB ARAP, HELZ, CGNL1, RDX, DENND5A, LY6E, RASSF2, SLA, FXYD5, USF3, TRIM28, IFITM1, SVIL, TRIOBP, MT-TI, RPS5, PRPF8, EEF2, CD164, FLOT1, HSPA8, HBA2, LTA4H, PNRC1, SHANK3, CD48, TNIP1, SBNO2, PHC2, ARRB2, MCTP2, and BCL2A1. In some embodiments, the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK, ANXA6, VCAN, PGD, PHACTR2, S100A8, PFN1, VIM, MIB1, TCF7, NRIP1, CSF3R, MX1, BIRC2, PTGES3, POU2F2, FKBP5, HDLBP, H2BC9, DARS1, AKNA, LCP1, SH3BP5, CCDC88C, DST, EIF4B, MVP, GRB2, HIP1, CD74, NFE2L1, WIPF1, PEAK1, CSK, TBC1D14, PTMA, KDM6B, LYZ, NUCKS1, RAPGEF1, MT-TL1, HIF1A, TXNIP, YBX1, GPI, PKP4, RPL21, RABEP1, TDP2, MAP4K4, GABARAP, HELZ, CGNL1, RDX, DENND5A, LY6E, and RASSF2. In some embodiments, the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK, ANXA6, VCAN, PGD, PHACTR2, S100A8, PFN1, VIM, MIB1, TCF7, NRIP1, CSF3R, MX1, BIRC2, PTGES3, POU2F2, FKBP5, HDLBP, H2BC9, DARS1, AKNA, LCP1, SH3BP5, CCDC88C, DST, EIF4B, MVP, GRB2, HIP1, and CD74. In some embodiments, the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK, ANXA6, VCAN, PGD, PHACTR2, andS100A8. In some embodiments, the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, and HEMGN. In some embodiments, the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, and CD44. In some embodiments, the genes in the panel of biomarkers differentiate between Kawasaki disease (KD), Multisystem Inflammatory Syndrome in Children (MIS-C), viral infections, and bacterial infections.

[0008] In some embodiments, the panel of biomarkers comprises one or more genes selected from histone protein coding genes and immune related genes. In some embodiments, the histone protein coding genes comprise one or more genes selected from the group consisting of Hl -10, Hl-2, Hl-3, Hl-4, Hl-5, H2AC11, H2AC12, H2AC13, H2AC14, H2AC16, H2AC17, H2AC20, H2AC21, H2AC4, H2AC6, H2AX, H2AZ1, H2AZ2, H2BC10, H2BC13, H2BC14, H2BC17, H2BC18, H2BC3, H2BU1, H3-3B, H3C1, H3C10, H3C11, H3C12, H3C2, H3C3, H3C7, H3C8, H4C1, H4C13, H4C2, H4C3, H4C4, H4C5, H4C6, and H4C9.

[0009] In some embodiments, the immune related genes comprise neutrophil markers. In some embodiments, the neutrophil markers comprise one or more genes selected from the group consisting of ABCA13, ACTR10, ADAM10, ALAD, ANO6, ANXA2, ARG1, ARL8A, ARMC8, ATP8A1, AZU1, B2M, BIN2, BPI, BRI3, C5AR1, CAMP, CANT1, CAT, CD 177, CD36, CD47, CD53, CD59, CDA, CEACAM1, CEACAM3, CEACAM6, CEACAM8, CKAP4, CLEC12A, CLEC4D, CLEC5A, CNN2, CPNE3, CTSG, CXCR1, CXCR2, CYBA, CYSTM1, DEFA1, DEFA4, DIAPH1, DNAJC13, DNAJC5, DYNLL1, ELANE, FABP5, FCAR, FCGR3A / FCGR3B, FPR1, FPR2, FTH1, GCA, GDI2, GMFG, GNS, GSTP1, GYG1, HEBP2, HMGB1, HP, HSP90AA1, IDH1, ILF2, IQGAP1, IQGAP2, ITGB2, LAMTOR2, LGALS3, LILRB2, LILRB3, LRG1, LTA4H, LTF, LYZ, MGAM, MGST1, MMP8, MMP9, MNDA, MPO, MS4A3, NME2, PA2G4, PADI2, PGLYRP1, PLAC8, PLAUR, PLD1, PPBP, PRDX6, PRTN3, PSMB7, PSMC2, PSMD3, PSMD7, PTPRJ, PYCARD, PYGB, QPCT, RAB10, RAB14, RAB27A, RAB31, RAB37, RAB3D, RAB6A, RAP1B, RAP2B, RETN, RHOG, RNASE2, S100A11, S100A12, S100A8, S100A9, SIOOP, SELL, SERPINA1, SERPINB10,SIRPA, SIRPB1 , SLC2A5, SLC44A2, SLC04C1, SLPI, SNAP23, SRP14, STOM, SVIP, TCN1, TIMP2, TMBIM1, TMEM63A, TOMI, TRAPPCI, TUBB, TYROBP, VCL, XRCC5, XRCC6, YPEL5, AKT3, ATP5F1B, ATP5PB, C1QB, CASP1, CASP4, CCL5, CCR1, CHP1, COL17A1, COL24A1, GP1BA, IFNAR1, IGHA1, IGHA2, IGHG1, IGKC, IGLC2, IL1B, ITGA2, ITGB1, ITPR1, ITPR2, JCHAIN, MCL1, MT-ATP6, MT-CYB, MT-ND1, MT-ND2, MT-ND3, MT-ND4, MT-ND4L, MT-ND5, MT-ND6, NCF2, NDUFA5, NDUFA6, NDUFB1, NDUFB2, NDUFB6, NDUFB9, NDUFS6, NLRP3, ORAI1, PF4, PIK3C3, PIK3CB, PLA2G12A, PLA2G4A, PLAAT1, PLCB4, PLCHI, PNPLA8, PPP3R1, RAC2, SDHB, SELPLG, TIMM13, TLR4, TSPO, UQCRFS1, VDAC1, VDAC3, VSTM1, ARPC2, ARPC3, CALM1, GNAI3, GNA01, GNAQ, GNAS, GNAZ, GNB1, GNB2, GNB5, GNG10, GNG11, GNG2, GNG5, KRAS, NFAT5, NFKBIA, NFKBIB, RALB, and RELB.

[0010] In some embodiments, the step of determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers comprises differentiating between Kawasaki disease (KD) and Multisystem Inflammatory Syndrome in Children (MIS-C) and the panel of biomarkers comprises TXNIP, ASH1L, YPEL5, PCBP1, STK17B, TRAK2, PTMA, RNA5SP149, AFF1, FHDC1, ANXA6, FKBP5, HSP90AB1, SYNE1, RBM33, VIM, PDCD4, CD44, RESF1, DAZAP2, COTL1, PFN1, EEF2, TRIR, and FTL. In some embodiments, the step of determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers comprises differentiating between Kawasaki disease (KD) and Multisystem Inflammatory Syndrome in Children (MIS-C) and the panel of biomarkers comprises EEF2 and / or FKBP5. In some embodiments, the panel of biomarkers comprises EEF2 and / or FKBP5 along with one or more additional genes. In some embodiments, the panel of biomarkers comprises EEF2 and / or FKBP5 along with 2-8 additional genes.

[0011] In some embodiments, the determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers comprises comparing the profile of the panel of biomarkers in the cell free RNA from the biological sample to a reference profile of the panel of biomarkers. In some embodiments, the determining an inflammatory syndrome status in the subject based on the profile of the panel of biomarkers comprises comparing the levels of RNAs corresponding to the biomarkers in the panel to levels of RNAs corresponding to the biomarkers in the reference profile. In some embodiments, the inflammatory syndrome comprises Kawasakidisease (KD), Multisystem Inflammatory Syndrome in Children (MIS-C), viral infections, and bacterial infections. In some embodiments, the determining the inflammatory syndrome status in the subject based on the profile of the panel of biomarkers comprises determining that the subject is more likely to have one inflammatory syndrome as compared to other inflammatory syndromes, e.g., the subject is more likely to have KD as compared to MIS-C, viral infections or bacterial infections.

[0012] In some embodiments, the method further comprises: d) determining the cell type of origins from cell-free RNA; e) determining tissue damage based on the cell type of origin of the cell-free RNA; and f) determining a clinical decision support tool using the determination of inflammatory disease from the panel of biomarkers and determination of tissue damage from the cell type of origin measurements.

[0013] In some embodiments, the step of determining the cell type of origin of cfRNA comprises performing nucleotide sequencing of the cell-free RNA. In some embodiments, the step of estimating the cfRNA cell types comprises using a reference RNA-seq data set with a deconvolution algorithm. In some embodiments, tissue damage is measured by comparing the cell type of origin measurements to a reference group. In some embodiments, the clinical decision support tool comprises the predication of inflammatory condition and measurements of tissue damage.

[0014] In some embodiments, the biological sample is a plasma sample. In some embodiments, the biological sample is a serum sample. In some embodiments, the subject is a mammal. In some embodiments, the subject is a human. In some embodiments, the subject is in a pediatric subpopulation of human.

[0015] In some embodiments, the profile of the panel of biomarkers comprises subject-specific gene expression of the biomarkers in the panel. In some embodiments, the profile of the panel of biomarkers comprises a score determined based on the levels of RNAs corresponding to the biomarkers in the panel using a trained classification model, and wherein the score correlates with an inflammatory syndrome status.

[0016] In some embodiments, the cell-free RNA is substantially free of contaminant RNA.

[0017] In some embodiments, in the inflammatory syndrome status determined is used in determining treatment for the subject.

[0018] In some embodiments, the test performance of the trained classification model is measured through one or more of accuracy, sensitivity, specificity, and / or area under the receiver operating characteristic curve (ROC-AUC).

[0019] In some embodiments, the inflammatory syndrome status is determined at a greater than 80%, 85%, 90%, 91%, 92%, 93%, 94%, or 95% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 80%, 85%, 90%, 91%, 92%, 93%, 94%, or 95% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, or 95% specificity.

[0020] In some embodiments, the trained classification model comprises one or more models selected from the group consisting of generalized linear models with Ridge and LASSO feature selection (GLMNETRIDGE and GLMNETLASSO), support vector machines with linear and radial basis function kernel (SVMLin and SVMRAD), random forest (RF), random forest ExtraTrees (EXTRATREES), neural networks (NNET), linear discriminant analysis (LDA), nearest shrunken centroids (PAM), C5.0 (C5), k-nearest neighbors (KNN), naive bayes (NB), CART (RPART), generalized linear model (GLM), and greedy forward search algorithm (GFS).

[0021] In some embodiments, the method further comprises a step of treating the subject based on the inflammatory syndrome status. In some embodiments, when the inflammatory syndrome status is determined to be Kawasaki disease (KD), the step of treating the subject based on the inflammatory syndrome comprises treating the subject with intravenous immunoglobulin (IVIG). In some embodiments, when the inflammatory syndrome status is determined to be Multisystem Inflammatory Syndrome in Children (MIS-C), the step of treating the subject based on the inflammatory syndrome comprises treating the subject with IVIG and a low to moderate dose glucocorticoid. In some embodiments, when the inflammatory syndrome status is determined to be a bacterial infection, the step of treating the subject based on the inflammatory syndrome comprises treating the subject with an antibiotic. In some embodiments, when the inflammatory syndrome status is determined to be a viral infection, the step of treating the subject based on the inflammatory syndrome comprises treating the subject with a normal treatment plan for such a virus.

[0022] In some embodiments, the subject is immunocompromised.BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawings will be provided by the office upon request and payment of the necessary fee.

[0024] FIG. 1A-E. Sample Overview. A) Sample counts and distribution of hospital of origin for each disease group. “Other” indicates other hospitalized pediatric controls. B) Age distribution, C) C-reactive protein (CRP) levels, D) sex distribution, and E) ICU status distribution for each sample group.

[0025] FIG. 2A-B. cfRNA Quality Control. A) Sample sequencing metrics, including all sequenced reads, 5 ’-3’ bias, number of feature counts, and intron to exon ratio for all samples split by hospital of origin. B) Principal component analysis on variance stabilization transformed counts of all samples using the top 500 most variable genes.

[0026] FIG. 3A-B. Shared cfRNA signatures among different inflammation syndromes. A) Differential abundance analysis using DESeq2 was performed between healthy controls and each other group individually. Vertical columns indicate the number of overlapping genes that are significantly differentially abundant between groups and controls (Benj ami ni -Hochberg (BH) adjusted p-values < 0.05). Dots below bars indicate the groups being intersected. Horizontal columns indicate the total number of DAGs between groups and controls. B) Significantly enriched pathways in the set of genes found to be differentially abundant between healthy controls and each disease group. Average p-value and fold change used for pathway analysis (Qiagen, IP A).

[0027] FIG. 4A-B. cfRNA molecules elevated in disease. A) Sum variance stabilization transformation counts of histone protein transcripts found to be differentially abundant in healthy and all other conditions. Counts grouped by the histone protein the differentially abundant cfRNA code for. B) Variance stabilization transformation counts of ELANE and MPO transcripts.

[0028] FIG. 5A-F. cfRNA distinguishes KD and MIS-C. A) Overview of sample set and modeling scheme. B) Volcano plot of differentially abundant transcripts between MIS-C andKD. Analysis was performed using the training data set (DESeq2). C) Adjusted p-value, base mean, and gene ROC AUC distributions for all significant genes from the training KD vs MIS-C comparison. The red dashed line represents the threshold cutoff used for filtering prior to model training. Green shaded area indicates genes used. D) ROC AUC values for the 14 machine learning classification models applied to training and validation sets. E) ROC-AUC curves of the training, validation, and test sets using the GLMNET with LASSO regression algorithm. F) Violin plots of the classifier scores from the GLMNET with LASSO regression algorithm.

[0029] FIG. 6A-C. Multiclass Classification of pediatric disease using cfRNA. A) Overview of machine learning framework used for multiclass classification. B) ROC-AUC plots for each one- vs-one model trained, along with train (top) and test (bottom) classifier score distributions. C) Confusion matrix of reference and predicted diagnoses for train and test samples. Color indicates fraction of samples in each category.

[0030] FIG. 7A-E. cfRNA as a clinical decision support tool. A) Median scaled cell type of origin fractions for samples separated by liver damage, cardiac function, COVID-19 severity, and endothelial damage, and healthy controls (Methods). Stars indicate statistically significant differences between groups in comparison (Wilcoxon rank sum test, BH adjusted p-value < 0.05). B-E) Case studies of patients with example clinical decision support tool results from the multiclass algorithm, along with measurements of endothelium, heart, liver, lung, and neuronal damage from the deconvolution data. Sample shown in red and healthy donor samples in gray. Z-scores calculated relative to healthy donor sample distributions. Endothelium refers to endothelial cell, heart to cardiac muscle cells, liver to hepatocyte, lung to club cell and type I pneumocyte, and neuronal to Schwann cell cfRNA cell type of origin fractions.

[0031] FIG. 8A-C. Characterization of disease. A) Cell type of origin fractions of skeletal and smooth muscle cells in patients with abnormal and normal cardiac function. Significance calculated with a Wilcoxon test and p-values adjusted using the BH method. B) ROC-AUC plot and C) classification scores of the train and test sets for classifying COVID- 19 vs other viral infections as modeled by a GLMNET LASSO.DETAILED DESCRIPTION

[0032] Disclosed herein are methods of plasma cell-free RNA (cfRNA) profiling by RNA sequencing to compare host immune and cellular injury responses associated with multiple inflammatory syndromes and / or infectious syndromes in children. Shared signatures of particular syndromes are identified across multiple conditions, highlighting the importance of incorporating multiple comparison groups for the development of disease-specific biomarkers. Using cfRNA profiles, the present disclosure shows a machine learning model with high accuracy in differentiating between Kawasaki disease (KD) and Multisystem Inflammatory Syndrome in Children (MIS-C), two pediatric inflammatory syndromes with overlapping clinical presentations. The disclosed methodology is expanded to construct a multi-class diagnostic classifier capable of distinguishing among KD, MIS-C, viral, and bacterial infections. In addition, the present disclosure demonstrates that the cfRNA profile can be correlated with markers of tissue damage and may be able to differentiate among different viral infections. The present disclosure presents an application of cfRNA profiling as a decision support tool for differential diagnosis of inflammatory syndromes in the clinical setting.

[0033] Although claimed subject matter will be described in terms of certain examples, other examples, including examples that do not provide all the benefits and features set forth herein, are also within the scope of this disclosure. Various structural, logical, and process step changes may be made without departing from the scope of the disclosure.

[0034] Ranges of values are disclosed herein. The ranges set out a lower limit value and an upper limit value. Unless otherwise stated, the ranges include the lower limit value, the upper limit value, and all values between the lower limit value and the upper limit value, including, but not limited to, all values to the magnitude of the smallest value (either the lower limit value or the upper limit value).

[0035] In the description that follows, certain conventions will be followed as regards to the usage of terminology. Generally, terms used herein are intended to be interpreted consistently with the meaning of those terms as they are known to those of skill in the art. In practicing the present disclosure, many conventional techniques in molecular biology, microbiology, cell biology, biochemistry, and immunology are used, which are within the skill of the art. These techniques are described in greater detail in, for example, Molecular Cloning: a LaboratoryManual 4th edition, I F. Sambrook and D.W. Russell, ed. Cold Spring Harbor Laboratory Press 2012; Recombinant Antibodies for Immunotherapy, Melvyn Little, ed. Cambridge University Press 2009; “Oligonucleotide Synthesis” (M. J. Gait, ed., 1984); “Animal Cell Culture” (R. I. Freshney, ed., 1987); “Methods in Enzymology” (Academic Press, Inc.); “Current Protocols in Molecular Biology” (F. M. Ausubel et al., eds., 1987, and periodic updates); “PCR: The Polymerase Chain Reaction”, (Mullis et al., ed., 1994); “A Practical Guide to Molecular Cloning” (Perbal Bernard V., 1988); “Phage Display: A Laboratory Manual” (Barbas et al., 2001). The contents of these references and other references containing standard protocols, widely known to and relied upon by those of skill in the art, including manufacturers’ instructions are hereby incorporated by reference as part of the disclosure.

[0036] The term “biological sample” includes body samples from an animal, including biological fluids such as serum, plasma, vitreous fluid, lymph fluid, synovial fluid, follicular fluid, seminal fluid, amniotic fluid, milk, whole blood, urine, cerebro-spinal fluid, saliva, sputum, tears, perspiration, mucus, and tissue culture medium, as well as tissue extracts such as homogenized tissue, and cellular extracts. In some embodiments, the biological sample is a serum, plasma or urine sample. In some embodiments, the biological sample is a plasma sample. In some embodiments, the biological sample is a serum sample.

[0037] The term “subject” refers to mammals. Non-limiting examples of mammals include but are not limited to human, horse, camel, dog, cat, pig, cow, goat, and sheep. In some embodiments, the mammal is human. In some embodiments, the subject is in the pediatric subpopulation of human, i.e. is a child. As used herein, a “pediatric subpopulation of human” or “child” refers to a human subject under 18 years of age.

[0038] The terms “cell-free RNA” or “cfRNA” refer to cfRNA released by cells in both the blood compartment and from vascularized solid tissues. cfRNA contains information about systemic immune dynamics and immune-tissue interactions. cfRNA is primarily released by dying cells; therefore, cfRNA may provide insights into pathways of cell death and mechanisms of cellular injury. cfRNA is also released into the blood by way of active secretion, cells release cfRNA to the subject’s bodily fluid, and thus may increase the quantity of the specific cfRNA in the subject’s biological sample as compared to a healthy individual. cfRNA may include anytypes of RNA that are circulating in the bodily fluid of a person without being enclosed in a cell body or a nucleus.

[0039] A “gene panel” or “panel” refers to a defined selection of genes that are enriched or sequenced. A panel will include relevant pathogen associated genes and likely variants of selected genes. In some embodiments, the panel comprises a collection of signature genes. As used herein, the term “signature gene” refers to a gene whose expression is correlated, either positively or negatively, with disease extent or outcome or with another predictor of disease extent or outcome. In some embodiment, the gene panels are used in sequencing assays. Sequencing based assays are known in the art. Some non-limiting examples of sequencing based assays are next generation sequencing (NGS), Sanger sequencing, oxidative bisulfite sequencing, direct RNA sequencing, and others. In a next generation sequencing (NGS) panel test, only clinically important genes are examined to obtain genomic data in a timely and cost-effective manner. In some embodiments, the gene panels are used in NGS. In some embodiments, the gene panels are used in non-sequencing based assays. Non-sequencing based assays are known in the art. Non-limiting examples of non-sequencing based assays include microarrays, polymerase chain reaction (PCR), including real-time PCR, reverse transcription PCR (RT- PCR), quantitative reverse transcription PCR (RT-qPCR), quantitative PCR (qPCR), digital droplet PCR (ddPCR), and others (as described in Shemer, R. et al., Current Protocols in Molecular Biology, 127.1 (2019): e90; and Zemmour, Hai, et al., Nature Communications, 9.1 (2018): 1-9, both incorporated herein in their entirety)..

[0040] As used herein, the term “profile” generally refers to gene expression profile. A gene expression profile can be understood to mean a pattern of abundance of expression of genes. In some embodiments, a gene expression profile is a measurement of the expression of multiple genes at once. In some embodiments, gene expression is measured by cfRNA in a subject, e.g., cfRNA in a serum or plasma sample, in which case, the profile is that of a subject’s cfRNA. Gene expression profiles can be characteristic or unique to a status of a subject, e.g., a healthy status or a disease status (such as cancer or infectious disease) and can therefore be used to distinguish between different statuses. In some embodiments, the profile is a profile of genes identified in this disclosure to be associated with infectious or inflammatory syndromes, which are also referred to herein as signature genes. It is to be understood that the profile of signaturegenes in this disclosure have not all been associated with the infectious or inflammatory syndromes previously. Signature genes of this disclosure were identified using the methods developed and disclosed herein. Signature genes exhibit differential expression in subjects having the infectious or inflammatory syndromes relative to subjects without infectious or inflammatory syndromes, for example. The differential expression refers to the difference in abundance of a signature gene in subjects having infectious or inflammatory syndromes relative to subjects without infectious or inflammatory syndromes. In some embodiments, the expression levels of signature genes may be used to predict progression of infectious or inflammatory syndromes. A “signature nucleic acid” is a nucleic acid comprising or corresponding to, in case of cDNA, the complete or partial sequence of a RNA transcript encoded by a signature gene, or the complement of such complete or partial sequence. A signature protein is encoded by or corresponding to a signature gene of the disclosure.

[0041] As used herein, the term “reference profile” refers to the profile of genes (e.g., the same genes identified in this disclosure to be associated with an infectious or inflammatory disease state) in a subject not inflicted with an infectious or inflammatory disease state.

[0042] In some embodiments, machine learning and model training is performed using R (v4.1.3) with the DESeq2 (vl.34.0), Caret (v6.0.90), and pROC (vl.18.0) packages. In some embodiments, sample metadata and count matrices split 70 / 30 into a training set and a test set, controlling for disease status, HIV status, and cohort to minimize differences in the training and test datasets. In some embodiments, sample metadata and count matrices are split into a training set, validation set, and test set, controlling for disease status, HIV status, and cohort to minimize differences in the datasets.

[0043] In some embodiments, features for model training are selected by filtering and differential abundance analysis. First, differential abundance analysis was performed on the raw training counts using DESeq2. Then feature selection was conducted using the output from DESeq2 on retained genes. In some embodiments, genes are excluded that have a base-mean of less than 100 and a Benjamini-Hochberg adjusted p-value greater than 0.05. In some embodiments, genes are excluded that have a base-mean of less than 50 and a Benjamini- Hochberg adjusted p-value greater than 0.05.

[0044] In some embodiments, machine learning algorithms are trained using 5-fold cross validation and grid search hyperparameter tuning. In some embodiments, accuracy, sensitivity, specificity, and / or area under the receiver operating characteristic curve (ROC-AUC) are used to measure test performance. In some embodiments, the classification models used are generalized linear models with Ridge and LASSO feature selection (GLMNETRIDGE and GLMNETLASSO), support vector machines with linear and radial basis function kernel (SVMLin and SVMRAD), random forest (RF), random forest ExtraTrees (EXTRATREES), neural networks (NNET), linear discriminant analysis (LDA), nearest shrunken centroids (PAM), C5.0 (C5), k-nearest neighbors (KNN), naive bayes (NB), CART (RPART), and generalized linear models (GLM).

[0045] As used herein, “differentiate” refers to the ability to assess disease severity and decipher between numerous afflictions. Differential diagnosis of pediatric inflammatory syndromes relies on clinical assessment of signs and symptoms and results from a broad array of laboratory tests. Culture-based and molecular assays are routinely used to identify viral and bacterial pathogens, but these tests do not interrogate the host response and hence are unable to differentiate between infectious and non-infectious conditions. In some embodiments, the present disclosure presents methods which assess disease severity and differentiate among inflammatory conditions including KD, MIS-C, viral, and bacterial infection.

[0046] The current disclosure is directed to cell-free RNA (cfRNA) profding to compare host immune and cellular injury responses associated with different inflammatory and / or infectious syndromes in children. The current disclosure describes methods of differentiating and detecting different inflammatory and / or infectious syndromes in children.In one aspect, the current disclosure is directed to a method comprising: a) obtaining cell-free RNA from a biological sample of a subject, wherein the biological sample is a plasma sample or a serum sample; b) determining a profile of a panel of biomarkers in the cell-free RNA; and c) determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers.

[0047] As used herein, “disease status” refers to inflammatory and / or infectious syndromes related conditions or progression of inflammatory and / or infectious syndromes. Gene expressionprofiles can be characteristic or unique to a status of a subject, e g., a healthy status or a disease status (such as cancer or infectious disease) and can therefore be used to distinguish between different statuses. In some embodiments, a disease status comprises disease free status, a latent disease status, an active disease status, or a progressing disease status. As used herein, “inflammatory disease status” refers to inflammatory related conditions or progression of inflammatory disease. As used herein, “infectious disease status” refers to infectious related conditions or progression of infectious disease. In some embodiments, the inflammatory syndrome comprises Kawasaki disease (KD). In some embodiments, the inflammatory syndrome comprises Multisystem Inflammatory Syndrome in Children (MIS-C). In some embodiments, the infectious disease comprises viral infections. In some embodiments, the infectious disease comprises bacterial infections.

[0048] As used herein, “determining a disease status” refers to the conclusion that a subject has the disease / syndrome. In some embodiments, determining disease status comprises differentiating a disease / syndrome from another disease or syndrome. For example, distinguishing KD from MIS-C, or vice versa. In some embodiments, determining disease status comprises differentiating a disease / syndrome from other diseases or syndromes. For example, distinguishing MIS-C from KD, viral infections, and bacterial infections, or distinguishing KD from MIS-C, viral infections, and bacterial infections. In some embodiments, the determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers comprises comparing the profile of the panel of biomarkers in the cell free RNA from the biological sample to a reference profile of the panel of biomarkers. In some embodiments, the determining an inflammatory syndrome status in the subject based on the profile of the panel of biomarkers comprises comparing the levels of RNAs corresponding to the biomarkers in the panel to levels of RNAs corresponding to the biomarkers in the reference profile. In some embodiments, the inflammatory syndrome comprises Kawasaki disease (KD), Multisystem Inflammatory Syndrome in Children (MIS-C), viral infections, and bacterial infections. In some embodiments, the determining the inflammatory syndrome status in the subject based on the profile of the panel of biomarkers comprises determining that the subject is more likely to have one inflammatory syndrome as compared to other inflammatory syndromes, e.g., the subject is more likely to have KD as compared to MIS-C, viral infections or bacterial infections.

[0049] In some embodiments, the step of determining a profile of a panel of biomarkers comprises performing nucleotide sequencing of the cell-free RNA. In some embodiments, the step of determining a profile of a panel of biomarkers comprises measuring the levels of RNAs corresponding to the biomarkers in the panel. In some embodiments, the panel of biomarkers comprises one, more, or all (including, e.g., at least 30, at least 50, at least 75, or at least 100) of genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK, ANXA6, VC AN, PGD, PHACTR2, S100A8, PFN1, VIM, MIB1, TCF7, NRIP1, CSF3R, MX1, BIRC2, PTGES3, POU2F2, FKBP5, HDLBP, H2BC9, DARS1, AKNA, LCP1, SH3BP5, CCDC88C, DST, EIF4B, MVP, GRB2, HIP1, CD74, NFE2L1, WIPF1, PEAK1, CSK, TBC1D14, PTMA, KDM6B, LYZ, NUCKS1, RAPGEF1, MT-TL1, HIF1A, TXNIP, YBX1, GPI, PKP4, RPL21, RABEP1, TDP2, MAP4K4, GAB ARAP, HELZ, CGNL1, RDX, DENND5A, LY6E, RASSF2, SLA, FXYD5, USF3, TRIM28, IFITM1, SVIL, TRIOBP, MT-TI, RPS5, PRPF8, EEF2, CD164, FLOT1, HSPA8, HBA2, LTA4H, PNRC1, SHANK3, CD48, TNIP1, SBNO2, PHC2, ARRB2, MCTP2, and BCL2A1.

[0050] In some embodiments, the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK, ANXA6, VC AN, PGD, PHACTR2, S100A8, PFN1, VIM, MIB1, TCF7, NRIP1, CSF3R, MX1, BIRC2, PTGES3, POU2F2, FKBP5, HDLBP, H2BC9, DARS1, AKNA, LCP1, SH3BP5, CCDC88C, DST, EIF4B, MVP, GRB2, HIP1, CD74, NFE2L1, WIPF1, PEAK1, CSK, TBC1D14, PTMA, KDM6B, LYZ, NUCKS1, RAPGEF1, MT-TL1, HIF1A, TXNIP, YBX1, GPI, PKP4, RPL21, RABEP1, TDP2, MAP4K4, GAB ARAP, HELZ, CGNL1, RDX, DENND5A, LY6E, and RASSF2.

[0051] In some embodiments, the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK,ANXA6, VC AN, PGD, PHACTR2, S100A8, PFN1, VIM, MIB1 , TCF7, NRIP1 , CSF3R, MX1, BIRC2, PTGES3, POU2F2, FKBP5, HDLBP, H2BC9, DARS1, AKNA, LCP1, SH3BP5, CCDC88C, DST, EIF4B, MVP, GRB2, HIP1, and CD74.

[0052] In some embodiments, the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK, ANXA6, VC AN, PGD, PHACTR2, and S100A8.

[0053] In some embodiments, the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, and HEMGN

[0054] In some embodiments, the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA1 1, BNIP3L, HIPK2, and CD44.

[0055] In some embodiments, the genes in the panel of biomarkers differentiate between Kawasaki disease (KD), Multisystem Inflammatory Syndrome in Children (MIS-C), viral infections, and bacterial infections.

[0056] In some embodiments, the panel of biomarkers comprises one or more genes selected from histone protein coding genes and immune related genes. In some embodiments, the histone protein coding genes comprise one or more genes selected from the group consisting of Hl-10, Hl-2, Hl-3, Hl-4, Hl-5, H2AC11, H2AC12, H2AC13, H2AC14, H2AC16, H2AC17, H2AC20, H2AC21, H2AC4, H2AC6, H2AX, H2AZ1, H2AZ2, H2BC10, H2BC13, H2BC14, H2BC17, H2BC18, H2BC3, H2BU1, H3-3B, H3C1, H3C10, H3C11, H3C12, H3C2, H3C3, H3C7, H3C8, H4C1, H4C13, H4C2, H4C3, H4C4, H4C5, H4C6, and H4C9.

[0057] In some embodiments, the immune related genes comprise neutrophil markers.Neutrophils can amount to as much as 70% of all leukocytes in the human body and are a major component of the immune system. Several neutrophil markers are commonly known and used in the art. Some common neutrophils are used to describe various neutrophil subpopulations depending on stages of neutrophil life cycle, including maturation, activation, migration, andvarious other neutrophil subsets, for example, Low-density neutrophils (LDN), PMN-MDSC (polymorphonuclear MD SC, also called granulocytic-MDSC), Proangiogenic neutrophils (PAN), and tumor-associated neutrophils (TAN). In some embodiments, the neutrophil markers comprise one or more genes selected from the group consisting of ABCA13, ACTR10, ADAM I 0, ALAD, ANO6, ANXA2, ARG1, ARL8A, ARMC8, ATP8A1, AZU1, B2M, BIN2, BPI, BRI3, C5AR1, CAMP, CANT1, CAT, CD177, CD36, CD47, CD53, CD59, CDA, CEACAM1, CEACAM3, CEACAM6, CEACAM8, CKAP4, CLEC12A, CLEC4D, CLEC5A, CNN2, CPNE3, CTSG, CXCR1, CXCR2, CYBA, CYSTM1, DEFA1, DEFA4, DIAPH1, DNAJC13, DNAJC5, DYNLL1, ELANE, FABP5, FCAR, FCGR3A / FCGR3B, FPR1, FPR2, FTH1, GCA, GDI2, GMFG, GNS, GSTP1, GYG1, HEBP2, HMGB1, HP, HSP90AA1, IDH1, ILF2, IQGAP1, IQGAP2, ITGB2, LAMTOR2, LGALS3, LILRB2, LILRB3, LRG1, LTA4H, LTF, LYZ, MGAM, MGST1, MMP8, MMP9, MNDA, MPO, MS4A3, NME2, PA2G4, PADI2, PGLYRP1, PLAC8, PLAUR, PLD1, PPBP, PRDX6, PRTN3, PSMB7, PSMC2, PSMD3, PSMD7, PTPRJ, PYCARD, PYGB, QPCT, RAB10, RAB14, RAB27A, RAB31, RAB37, RAB3D, RAB6A, RAP IB, RAP2B, RETN, RHOG, RNASE2, S100A11, S100A12, S100A8, S100A9, SI OOP, SELL, SERPINA1, SERPINB10, SIRPA, SIRPB1, SLC2A5, SLC44A2, SLCO4C1, SLPI, SNAP23, SRP14, STOM, SVIP, TCN1, TIMP2, TMBIM1, TMEM63A, TOMI, TRAPPCI, TUBB, TYROBP, VCL, XRCC5, XRCC6, YPEL5, AKT3, ATP5F1B, ATP5PB, C1QB, CASP1, CASP4, CCL5, CCR1, CHP1, COL17A1, COL24A1, GP1BA, IFNAR1, IGHA1, IGHA2, IGHG1, IGKC, IGLC2, IL1B, ITGA2, ITGB1, ITPR1, ITPR2, JCHAIN, MCL1, MT-ATP6, MT-CYB, MT-ND1, MT-ND2, MT-ND3, MT-ND4, MT-ND4L, MT-ND5, MT-ND6, NCF2, NDUFA5, NDUFA6, NDUFB I , NDUFB2, NDUFB6, NDUFB9, NDUFS6, NLRP3, ORAI1, PF4, PIK3C3, PIK3CB, PLA2G12A, PLA2G4A, PLAAT1, PLCB4, PLCHI, PNPLA8, PPP3R1, RAC2, SDHB, SELPLG, TIMM13, TLR4, TSPO, UQCRFS1, VDAC1, VDAC3, VSTM1, ARPC2, ARPC3, CALM1, GNAI3, GNAO1, GNAQ, GNAS, GNAZ, GNB1, GNB2, GNB5, GNG10, GNG11, GNG2, GNG5, KRAS, NFAT5, NFKBIA, NFKBIB, RALB, and RELB.

[0058] In some embodiments, the step of determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers comprises differentiating between Kawasaki disease (KD) and Multisystem Inflammatory Syndrome in Children (MIS-C) and thepanel of biomarkers comprises TXNIP, ASH1L, YPEL5, PCBP1, STK17B, TRAK2, PTMA, RNA5SP149, AFF1, FHDC1, ANXA6, FKBP5, HSP90AB1, SYNE1, RBM33, VIM, PDCD4, CD44, RESF1, DAZAP2, COTL1, PFN1, EEF2, TRIR, and FTL. In some embodiments, the step of determining an inflammatory disease status in the subject based on the profde of the panel of biomarkers comprises differentiating between Kawasaki disease (KD) and Multisystem Inflammatory Syndrome in Children (MIS-C) and the panel of biomarkers comprises EEF2 and / or FKBP5. In some embodiments, the panel of biomarkers comprises EEF2 and / or FKBP5 along with one or more additional genes. In some embodiments, the panel of biomarkers comprises EEF2 and / or FKBP5 along with 2-8 additional genes.

[0059] In some embodiments, the determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers comprises comparing the profile of the panel of biomarkers in the cell free RNA from the biological sample to a reference profile of the panel of biomarkers. In some embodiments, the determining an inflammatory syndrome status in the subject based on the profile of the panel of biomarkers comprises comparing the levels of RNAs corresponding to the biomarkers in the panel to levels of RNAs corresponding to the biomarkers in the reference profile. In some embodiments, the inflammatory syndrome comprises Kawasaki disease (KD), Multisystem Inflammatory Syndrome in Children (MIS-C), viral infections, and bacterial infections. In some embodiments, the determining the inflammatory syndrome status in the subject based on the profile of the panel of biomarkers comprises determining that the subject is more likely to have one inflammatory syndrome as compared to other inflammatory syndromes, e.g., the subject is more likely to have KD as compared to MIS-C, viral infections or bacterial infections.

[0060] In some embodiments, the method further comprises: d) determining the cell type of origins from cell-free RNA; e) determining tissue damage based on the cell type of origin of the cell-free RNA; and f) determining a clinical decision support tool using the determination of inflammatory disease from the panel of biomarkers and determination of tissue damage from the cell type of origin measurements.

[0061] In some embodiments, the step of determining the cell type of origin of cfRNA comprises performing nucleotide sequencing of the cell-free RNA. In some embodiments, the step of estimating the cfRNA cell types comprises using a reference RNA-seq data set with a deconvolution algorithm. Deconvolution algorithms are known in the art, for example, the BayesPrism and the Tabula Sapiens human cell atlas as a reference. In some embodiments, tissue damage is measured by comparing the cell type of origin measurements to a reference group. The CTO estimates are compared with known biomarkers and other indicators of specific organ injury or dysfunction. In some embodiments, the clinical decision support tool comprises the predication of inflammatory condition and measurements of tissue damage.

[0062] In some embodiments, the biological sample is a plasma sample. In some embodiments, the biological sample is a serum sample. In some embodiments, the subject is a mammal. In some embodiments, the subject is a human. In some embodiments, the subject is in a pediatric subpopulation of human.

[0063] In some embodiments, the profile of the panel of biomarkers comprises subject-specific gene expression of the biomarkers in the panel. In some embodiments, the profile of the panel of biomarkers comprises a score determined based on the levels of RNAs corresponding to the biomarkers in the panel using a trained classification model, and wherein the score correlates with an inflammatory syndrome status.

[0064] In some embodiments, the cell-free RNA is substantially free of contaminant RNA. “Contaminant RNA” refers to RNA that is not from the subject. The contaminant RNA could be RNA from other bacteria in the subject. cfRNA that is substantially free of contaminant RNA means that the signature transcripts are from the subject or host RNA, rather than from RNA produced by a foreign subject.

[0065] In some embodiments, in the inflammatory syndrome status determined is used in determining treatment for the subject. In some embodiments, the inflammatory syndrome status determined is used to monitor treatment of the subject. In some embodiments, the inflammatory syndrome status determined is used to monitor efficacy of a treatment administered to the subject.

[0066] In some embodiments, the test performance of the trained classification model is measured through one or more of accuracy, sensitivity, specificity, and / or area under the receiver operating characteristic curve (ROC-AUC).

[0067] As used herein, “test accuracy” or “accuracy” provides the final generalization power. In some embodiments, the inflammatory syndrome status is determined at a greater than 80% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 85% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 86% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 87% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 88% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 89% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 90% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 91% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 92% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 93% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 94% accuracy. In some embodiments, the inflammatory syndrome status is determined at a greater than 95% accuracy.

[0068] As used herein, “sensitivity” describes the probability of a positive test result, conditioned on the individual truly being positive. Sensitivity is determined by the number of true positives divided by the sum of true positives and false negatives. As such, a test which reliably detects the presence of a condition, resulting in a high number of true positives and low number of false negatives, will have a high sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 80% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 85% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 86% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 87% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 88% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 89% sensitivity. In some embodiments, the inflammatory syndrome status isdetermined at a greater than 90% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 91% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 92% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 93% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 94% sensitivity. In some embodiments, the inflammatory syndrome status is determined at a greater than 95% sensitivity.

[0069] As used herein, “specificity” refers to the probability of a negative test result, conditioned on the individual truly being negative. Specificity is determined by the number of true negatives divided by the sum of true negatives and false positives. As such, a test which reliably excludes individuals who do not have the condition, resulting in a high number of true negatives and low number of false positives, will have a high specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 70% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 75% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 80% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 81% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 82% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 83% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 84% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 85% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 86% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 87% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 88% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 89% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 90% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 91% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 92% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 93% specificity. In some embodiments, the inflammatorysyndrome status is determined at a greater than 94% specificity. In some embodiments, the inflammatory syndrome status is determined at a greater than 95% specificity.

[0070] In some embodiments, the trained classification model comprises one or more models selected from the group consisting of generalized linear models with Ridge and LASSO feature selection (GLMNETRIDGE and GLMNETLASSO), support vector machines with linear and radial basis function kernel (SVMLin and SVMRAD), random forest (RF), random forest ExtraTrees (EXTRATREES), neural networks (NNET), linear discriminant analysis (LDA), nearest shrunken centroids (PAM), C5.0 (C5), k-nearest neighbors (KNN), naive bayes (NB), CART (RPART), generalized linear model (GLM), and greedy forward search algorithm (GFS).

[0071] In some embodiments, the method further comprises a step of treating the subject based on the inflammatory syndrome status. Treatments for Kawasaki disease (KD), Multisystem Inflammatory Syndrome in Children (MIS-C), bacterial infections, and viral infections are known in the art. In some embodiments, when the inflammatory syndrome status is determined to be Kawasaki disease (KD), the step of treating the subject based on the inflammatory syndrome comprises treating the subject with intravenous immunoglobulin (IVIG). In some embodiments, when the inflammatory syndrome status is determined to be Multisystem Inflammatory Syndrome in Children (MIS-C), the step of treating the subject based on the inflammatory syndrome comprises treating the subject with IVIG and a low to moderate dose glucocorticoid. In some embodiments, when the inflammatory syndrome status is determined to be a bacterial infection, the step of treating the subject based on the inflammatory syndrome comprises treating the subject with an antibiotic. In some embodiments, when the inflammatory syndrome status is determined to be a viral infection, the step of treating the subject based on the inflammatory syndrome comprises treating the subject with a normal treatment plan for such a virus.

[0072] In some embodiments, the subject is immunocompromised or immunosuppressed. As used herein, “immunocompromised or immunosuppressed” refers to having a weakened immune system. Subjects that are immunocompromised have a reduced ability to fight infections and other diseases. This may be caused by certain diseases or conditions or certain medicines or treatments.EXAMPLESExample 1. Clinical cohort.

[0073] 370 plasma samples were collected and analyzed from pediatric patients with inflammatory and infectious conditions at four hospitals in the US, including Rad Children’s Hospital San Diego (UCSD), Emory, Children’s National Hospital (CNH), and University of California San Francisco (UCSF) (FIG. 1A). This sample set included patients diagnosed with KD, MIS-C, viral infection, bacterial infection, and other hospitalized pediatric controls, some but not all with inflammatory disease (for example, arthralgia, Crohn’s disease flare, parenchymal lung disease, chronic lung disease, toxic shock syndrome, and post-vaccine myocarditis), as well as healthy children (FIG. 1A and Table 1). Included cases of bacterial and viral infections were heterogeneous with respect to infection site and pathogen. Age, sex, race, severity, and inflammatory status were recorded for each patient (FIGS. 1B-E). All patient samples were collected within 4 days of hospitalization during the acute phase of disease. RNA next-generation sequencing was used to quantify the genes and cell types of origin of cfRNA in each sample, with an average of 20.7 million reads sequenced per sample. No batch effect was observed between samples from different hospitals (FIG. 2A-B, Methods). Machine learning models were built to identify biosignatures associated with different diseases.Example 2. cfRNA signatures of disease.

[0074] The initial focus was on characterizing changes in cfRNA profiles in plasma that are common across the various inflammatory conditions. Pairwise differential gene abundance analysis was performed between the healthy control group and each disease group (Methods, FIG. 3A). This analysis identified differentially abundant genes (DAGs) for all comparisons (BH adjusted p-value < 0.05). The smallest number of DAGs (n=2,686) was identified when comparing healthy controls to patients in the other hospitalized control group, likely due to the smaller sample size and higher heterogeneity of this sample group. The largest number of DAGs (n=6,606) was identified when comparing healthy controls to patients with KD (FIG. 3A). Further analysis of the DAGs for all conditions revealed a significant number of shared genes (1,871 DAGs, FIG. 3A), including histone protein coding genes that were elevated for all disease groups, indicating that an increase in histone transcripts is a universal indicator of inflammation (FIG. 4A). The analysis also identified immune related transcripts MPO, ELANE,CD53, and CXCR2 elevated in each disease group, homeostasis related transcripts CDK19, ANAPC5, and 32 mitochondrial protein coding RNAs in the control group (FIG. 4B). Pathway analysis of the 1,871 overlapping DAGs revealed an enrichment of neutrophil and cell replication transcripts related to inflammation (FIG. 3B). These findings point to shared signatures of inflammation among disease groups and emphasize the need for inter-group comparisons to develop disease-specific biosignatures.Table 1 Sample Cohort OverviewExample 3. cfRNA can differentiate MIS-C and Kawasaki Disease.

[0075] KD and MIS-C share many clinical characteristics: they are highly inflammatory, present with endothelial dysfunction, and have multiple overlapping signs, including skin rashes, mucosal involvement, and fever. There are currently no molecular tests to distinguish between KD and MIS-C. Therefore, cfRNA was assessed for use to differentiate these severe pediatric inflammatory syndromes. The KD (n=101) and MIS-C (n=97) samples were divided into training (60%), validation (20%), and test (20%) sets, ensuring a roughly equal representation of hospital of origin and disease subclassification across all 3 sets (FIG. 5A). The training data was used for feature selection and to train machine learning models. The validation set was used to select the final model, based on the model with the highest receiving operator characteristic Area Under the Curve (AUC). Last, the performance of the final model was evaluated using the test set. To prevent influence of the test set on the training data and to ensure unbiased results, each set was normalized separately using a variance stabilizing transformation (Methods).

[0076] The initial analysis using the training set identified 1,242 differentially abundant genes between KD and MIS-C (DESeq2, Benjamini -Hochberg (BH) adjusted p-value < 0.01 and base mean > 10; FIG. 5B). This gene list was refined based on adjusted p-value, base mean, individual gene AUC, and fold change, resulting in a final tally of 132 genes for model input (FIG. 5C). 14 machine learning classification models were then trained, evaluating their performance on the validation set (Methods). The GLMNET model with LASSO regression exhibited the highest validation set AUC and was selected as our final model (FIG. 5D). A unique feature of the GLMNET LASSO algorithm is the feature selection step, which selected 25 genes for the modeling (Table 2). Using this trained model, the classification performance was tested using the test set (ROC-AUC train=1.00, validation=0.98, test=0.98) (FIG. 5E). Similar distributions were also observed in the classification scores across sample sets, further confirming that there was little to no overfitting of the training and test sets (FIG. 5F).

[0077] The KD vs MIS-C gene panel was compared to previously reported differentially expressed genes (DEGs) from a study of cultured endothelial cells treated with KD and MIS-C sera. Significant overlap in genes was found; nine of the twenty-five genes in our panel were also identified as DEGs in the previously reported study (nominal p-value < 0.05). Of the overlapping genes, those elevated in acute KD are associated with autophagy (ANXA6, CD44, PDCD4) and endothelial-mesenchymal transition (EndoMT) (STK17B, VIM), processes linked to endothelial cell function in inflammatory conditions. This provides evidence that the key differences in endothelial dysfunction and cardiac outcomes between MIS-C and KD may be linked to differences in endothelial autophagy and activation of EndoMT pathways.Table 2. Gene panel for KD vs MIS-C predictionExample 4. Multi-disease classification using cfRNA.

[0078] Then, it was determined if cfRNA could be used in the more challenging scenario of multi-disease classification. For this, a machine learning framework was developed that combines the outputs of one-vs-one models using a random forest multiclass algorithm (Methods, FIG. 6A). The dataset was first split into training (70%) and test (30%) sets, with roughly comparable proportions of samples from patients with KD, MIS-C, viral infection, or bacterial infection, while also stratifying the groups evenly with respect to hospital of origin and disease subclassification. Differential abundance analysis was performed for each pairwise comparison using only the training data (Dataset S5). Next, individual one-vs-one GLMNET models with LASSO regression were trained for each sample group (MIS-C, KD, viral infection, or bacterial infection), using the top 100 genes identified in the differential abundance analysis. These genes were selected based on gene abundance AUC and level of significance (Methods). Each of the one-vs-one models demonstrated high performance with individual model scores classifying samples with high accuracy (Test AUC: min=0.86, max=0.99; Test Accuracy: min=0.83, max=0.92; FIG. 6B, Table 3). The union of genes used by each GLMNET model with LASSO regression generated a 109-gene panel. To combine the outputs of the individual models, a multiclass random forest classifier was trained using the classification scores from the one-vs-one models. Using this framework, the multiclass machine learning model achieved high accuracy in both the training and test sets (Accuracy, train=100% and test=80%). The bacterial infection group had the highest rate of misclassified samples, likely due to the smaller sample size and heterogeneity of both pathogen and site of infection (Table 3). The high performanceof the multiclass machine learning model on a relatively small number of genes points to the potential utility of cfRNA in differentiating complex inflammatory conditions in a clinical setting.Table 3. Model performance metricsExample 5. Characterization of disease using cfRNA.

[0079] It was determined if cfRNA could be employed not only for classification but also for disease characterization. Since inflammation and / or infection can impact multiple organsystems, understanding organ-specific damage and function is crucial for guiding clinical management. Previous work from our group and others has shown that quantifying the cfRNA cell-type-of-origin (CTO) is a viable non-invasive method to assess cell, tissue and organspecific injury. Here, deconvolution of the cell types of origin of cfRNA was employed by using BayesPrism and the Tabula Sapiens human cell atlas as a reference. The CTO estimates were compared with known biomarkers and other indicators of specific organ injury or dysfunction (Methods, FIG. 7A). Hepatocyte and intrahepatic cholangiocyte contributions were first compared to the cfRNA in plasma to alanine transaminase measurements (ALT, n=141). Significantly elevated levels of cfRNA were observed from hepatocytes and intrahepatic cholangiocytes in patients with high ALT (ALT > 100 IU / L) compared to patients with normal ALT (ALT < 40 IU / L, FIG. 7A). To explore cardiac function and damage, samples were categorized as having either normal or abnormal cardiac function (Methods). In the abnormal group, increased levels of cfRNA were observed from cardiac muscle cells and pericytes, likely indicative of increased cardiac cell injury or death. Interestingly, this group also exhibited elevated cfRNA levels from kidney epithelial cells, intrahepatic cholangiocytes, club cells and type I pneumocytes, and bronchial epithelium, suggesting other end organ damage associated with impaired cardiac function (FIG. 7A). Patients with abnormal cardiac function did not have elevated levels of smooth or skeletal muscle cell derived cfRNA (FIG. 8A). Last, to evaluate lung function and damage samples from healthy individuals and viral infection patients with COVID- 19 were tested, further stratifying the COVID-19 samples by disease severity (Methods). Elevated levels of cfRNA were observed from club cells and type I pneumocytes in moderate COVID-19 cases compared to healthy individuals, with a greater increase for severe cases (FIG. 7A). Similar trends were observed in the levels of bronchial epithelium derived cfRNA; however, the difference was not statistically significant.

[0080] Next, it was assessed if cfRNA could be used to further stratify sample groups by distinguishing between COVID- 19 infection from SARS-CoV-2 and infections from other viral pathogens (Influenza, RSV, EBV, Adenovirus, etc). Viral infection samples were separated into COVID- 19 and non-COVID-19 viral infections and randomly split the data into training (70%) and testing (30%) data sets. Next, a GLMNET with LASSO regression model was trained to differentiate between these groups (Methods). The trained model had high performance on thetraining and test sets despite including only seven genes (train AUC = 0.99, test AUC = 0.93), demonstrating the potential of cfRNA for respiratory viral species differentiation (FIG. 8B-C). Example 6. cfRNA profiling as a clinical decision support tool.

[0081] A cfRNA “report” was then simulated for each patient in the multiclass test set and discuss four patient cases in detail to illustrate how cfRNA can be integrated to support clinical decision making. Each report includes diagnostic predictions from the multiclass classification algorithm and predicted organ involvement levels as z-scores of CTO fractions in samples from patients with disease compared to healthy patients.

[0082] The first case is a 13-year-old male with very-high-risk B-cell acute lymphoblastic leukemia in delayed intensification chemotherapy who presented with a 1-day history of fever (FIG. 7B). His laboratory assessments were notable for leukopenia, anemia, thrombocytopenia, elevated C-reactive protein (CRP), mildly elevated alanine transaminase, normal kidney function, and developed hypotension following admission. He had known household contacts with COVID- 19, and his nasopharyngeal PCR was positive for SARS-CoV-2, but he did not develop respiratory symptoms. The final diagnosis was COVID-19 in an immunocompromised host. Despite the immunocompromised state of the patient and lack of respiratory symptoms common in COVID-19, the cfRNA algorithm correctly identified the patient as having a viral infection. Furthermore, the elevated levels of liver and heart derived cfRNA are interesting given the clinical manifestations of mildly elevated ALT and hypotension.

[0083] In the second case, the cfRNA model correctly predicted the patient as having MIS-C, but the classifier score for MIS-C and viral infection were very comparable, producing what could be considered a “borderline” result (FIG. 7C). The patient is a 4-year-old male who presented with a 1-week history of abdominal pain, vomiting, and watery diarrhea, with subsequent development of fever, conjunctival injection and swelling of the hands and feet. The patient had normal white blood cell counts (WBC), with lymphopenia, thrombocytopenia, hyponatremia, acidosis, and elevated brain natriuretic peptide (BNP), C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), with normal liver and kidney function. He was SARS-CoV-2 PCR negative but SARS-CoV-2 immunoglobulin G positive. No additional viral testing was performed. His course was complicated by hypotension, but an echocardiogram was grossly normal. The final diagnosis was MIS-C, which was the top prediction from the cfRNA model.Interestingly, viral infection was the second most probable diagnosis from the cfRNA model. MIS-C is derived from a viral infection (SARS-CoV-2), and MIS-C and viral infection can be difficult to differentiate. There are at least 3 possible interpretations for these results: (1) the model is correctly identifying MIS-C along with a viral signature from SARS-CoV-2, (2) the patient is misdiagnosed with MIS-C and has only a viral infection, or (3) the detection of a viral signature is due to errors in the model. Elevated liver cfRNA was also observed despite the patient having normal liver enzymes along with elevated heart tissue derived cfRNA, which is consistent with the elevated BNP.

[0084] In the third case, the model incorrectly predicted the diagnosis in an immunocompromised patient, but the correct diagnosis of bacterial infection was the second most likely prediction (FIG. 7D). The patient is a 16-month-old male with a history of Wiskott- Aldrich syndrome requiring a bone marrow transplant. He was admitted with fever, fatigue, pancytopenia with severe thrombocytopenia, elevated CRP, and blood cultures positive for Klebsiella pneumoniae . The source of the infection is unknown but was believed to be secondary to bacterial gut translocation or central line infection. His pancytopenia and thrombocytopenia were attributed to his history of Wiskott-Aldrich syndrome. He was treated with cefepime and responded clinically to a full course of antibiotics. The final diagnosis was bacterial infection, which was the second ranked prediction by the cfRNA model. Interestingly, the patient presented with some symptoms characteristic of MIS-C, the apparently erroneous top prediction made by the cfRNA model; specifically with severe thrombocytopenia and elevated CRP, which are part of the MIS-C case definition.

[0085] In the last case, the model predicted an incorrect diagnosis, but the results of the multiclass model would have been informative in providing the correct diagnosis (FIG. 7E). The patient presented with fever during the COVID-19 pandemic and was initially diagnosed with MIS-C. However, a family member of the patient had recently been diagnosed with KD. Given the known genetic predisposition associated with KD, a re-evaluation of the original case was conducted. The patient was subsequently found to meet the American Heart Association’s case definition for KD and had initial laboratory values and a subsequent clinical course that were consistent with KD (including periungual desquamation). The patient was reclassified accordingly. Although the multi class model did not predict the proper diagnosis of KD, theclassifier scores for KD, bacterial infection, and viral infection were very similar, while the classifier score for MIS-C was much lower. Furthermore, the MIS-C vs KD pairwise classifier accurately predicted KD. While the final diagnosis of the multiclass model was incorrect, the multiclass model and MIS-C vs KD classifier results could have aided the clinicians in determining that the patient did not have MIS-C upon initial admission.

[0086] The data from this study provides opportunities to better understand cfRNA profiles in health and disease. cfRNA signatures were observed that are common to all inflammatory conditions, including elevated histone related RNAs, neutrophil extracellular trap (NET) components MPO and ELANE, as well as immune markers ILF2, IFI16, CD53, and CXCR2. NETs are composed of DNA, histones, and other proteins, and act to trap and neutralize pathogens while minimizing host cell damage. These observation of cfRNA signatures of NET formation is consistent with the reported role of NETs in KD and MIS-C.

[0087] To test the ability of cfRNA profiles to distinguish among pediatric inflammatory syndromes, cfRNA profiles were compared for patients diagnosed with KD and MIS-C, conditions that are very similar clinically. There is a need for improved molecular tools to discriminate between these two conditions as exemplified by a recent study from Day-Lewis et al. which reported significant overlap in signs and symptoms between KD and MIS-C (based on the 2023 case definition), with an estimated false positive rate of 8%. The cfRNA signature presented here has near-perfect accuracy and therefore has high potential for translation into a useful clinical molecular test. Furthermore, the identified gene signature provides new mechanistic insight into KD and MIS-C. For example, two genes included in the machine learning signature are EEF2, elevated in KD, and FKBP5, elevated in MIS-C. EEF2 is an elongation factor that has previously been implicated in senescence and exposure to bacterial toxins. This is compelling given that it has been suggested that an inhaled toxin or pathogen may be one of the triggers of KD. FKBP5 is a member of the immunophilin family and has been implicated in immune-stress response and other cellular processes in the brain and peripheral nervous system. This observation is relevant given the neuronal involvement reported for MIS- C. Beyond KD and MIS-C, cfRNA allows one to differentiate between CO VID-19 and other viral infections, opening the door for future studies to create more granular classificationalgorithms that can differentiate among pathogens at the species level based on cfRNA host profding.

[0088] This disclosure presents the first multiclass model for differential diagnosis of inflammatory syndromes using plasma cfRNA. The final model includes just 109 genes, suggesting that translating the model to a multiplexed qRT-PCR based platform with a rapid turnaround time of a few hours is likely practical. Furthermore, one of the most compelling aspects of cfRNA profiling is the ability to quantify the extent of organ involvement. We demonstrate this concept with simultaneous quantification of injury to multiple organ systems (liver, cardiac, lung, endothelium). Available clinical tests for evaluating tissue injury (for example, alanine aminotransferase levels in the liver) enabled us to confirm the accuracy and cfRNA-based detection of organ injury. Even more compelling is the potential clinical utility of cfRNA in quantifying injury to tissues where current tests are not available or lack adequate sensitivity, such as endothelial and neuronal injury.Example 7. Materials and General Methods.

[0089] Ethics Statement. The University of California, San Francisco (UCSF) Institutional Review Board (IRB) (#21-33403), San Francisco, CA; Emory University IRB (STUDY00000723), Atlanta, GA; Children’s National Medical Center IRB (ProOOO 10632), Washington, DC; and Cornell University IRB for Human Participants (2012010003), New York, NY each approved the protocols for this study. All samples and patient information were deidentified for analysis and shared with collaborating institutions. At Emory University, the IRB approved protocol was a prospective enrollment study under which parents provided consent and children assent as appropriate for age. At Children’s National Medical Center and UCSF, the IRB protocols were “no subject contact” sample biobanking protocols under which consent was not obtained and data was extracted from medical charts. At University of California, San Diego (UCSD), the IRB reviewed and approved collection and sharing of samples and data (IRB #140220). Signed consent and assent were provided by the parent(s) and pediatric patient, respectively.

[0090] Sample Acquisition. Samples were acquired from UCSF as previously described in Cell Rep. Med. 4, 101034 (2023). Briefly, hospitalized pediatric patients were identified as having COVID-19 by testing positive with SARS-CoV-2 real-time PCR (RT-PCR). Residual wholeblood samples were collected in EDTA lavender top tubes and diluted 1 : 1 : in DNA / RNA shield (Zymo Research). The remaining blood was centrifuged at 2500 rpm for 15 min and the available plasma was retained. All samples were stored at -80°C freezer until used. Samples were acquired from Emory and Children’s Healthcare of Atlanta as previously described Cell Rep. Med. 4, 101034 (2023). Briefly, pediatric patients were classified as having COVID-19 by SARS-CoV-2 RT-PCR and as having MIS-C if they met the CDC case definition. Controls were healthy outpatients with no known history of C OVID- 19 who volunteered for specimen collection. Whole blood was collected in EDTA lavender top tubes and aliquoted for plasma extraction via centrifugation at 2500 rpm for 15 min. Samples were stored at -80°C and shipped on dry ice to either UCSF or Cornell for analysis. Samples were acquired from UCSD prior to any treatment in all subjects in EDTA lavender top tubes and centrifuged at 2,000g for 10 minutes. Plasma was collected and stored at -80C. Samples were acquired from Children’s National Hospital as previously described. Briefly, pediatric patients were classified as having MIS-C if they met the CDC case definition. Whole blood samples were collected and centrifuged at 1300 x G for 5 minutes at room temperature. Plasma was aliquoted into a cryovial and frozen at -80°C.

[0091] Clinical Data. Patients were stratified as previously described Cell Rep. Med. 4, 101034 (2023). For the purposes of this study, patients were classified as having MIS-C by multidisciplinary teams that adjudicated whether a patient met the CDC case definition of MIS- C. COVID-19 was defined as any patient with PCR-confirmed SARS-CoV-2 infection within the preceding 14 days who did not also meet the MIS-C case definition. Kawasaki Disease patients at UCSD met the AHA definition for complete or incomplete KD. Viral and bacterial infection patients enrolled at UCSD were adjudicated and a final diagnosis assigned by the research team (one ED clinician and one pediatric infectious disease expert) 2-3 months after the acute illness to allow time for serologies, recurrence, and clinical recovery to be assessed. Patients with a self-limited illness that resolved without treatment and for whom viral studies were either negative or not done were classified as having a “viral syndrome”. Clinical data was abstracted from the medical record and submitted into a REDCap databases housed at UCSF or UCSD.

[0092] Sample processing and sequencing. Briefly, samples were received on dry ice, RNA was extracted, and libraries prepared and sequenced on a NextSeq or NovaSeq Illuminasequencer. Reads were trimmed to 61 bp and sequencing data was processed using a custom bioinformatics pipeline which included quality filtering and trimming, alignment to the human GRCh38 reference genome, and counting of gene features.

[0093] Sample quality filtering. Using the sequencing data, quality control was performed by analyzing DNA contamination, rRNA contamination, total counts, and RNA degradation. DNA contamination was estimated by calculating the ratio of reads mapping to introns and exons. rRNA contamination was measured using SAMtools (vl.14). Total counts were calculated using featureCounts (v2.0.0). Degradation was estimated by calculating the 5’-3’ bias using Qualimap (v2.2.1). Samples were removed from analysis if either the intron to exon ratio was greater than 3, if a sample had less than 75,000 total feature counts, or if the 5’-3’ read alignment ratio bias was greater than 2.

[0094] Differential abundance analysis. Gene transcript abundances were compared using a negative binomial model implemented using the DESeq2 R package. Gene transcript base mean abundance, adjusted p-value, and log2 fold change were taken from the DESeq2 Results output. Gene transcript AUC was calculated using VST transformed counts and the pROC R package.

[0095] Sample partitioning. Samples were partitioned for machine learning applications taking into consideration diagnosis, hospital of origin, and disease subclassification. MIS-C and COVID-19 samples were subclassified by severity, as previously defined. KD samples were subclassified by phenotypic subclusters, as previously defined. Non-COVID-19 Viral and bacterial infection samples were evenly partitioned by diagnosis and hospital of origin only.

[0096] Machine learning: MIS-C vs KD. Samples were partitioned into training, validation, and test sets at a ratio of 60:20:20, partitioning evenly based on factors such as diagnosis, severity / sub group, and hospital of origin. Subsequently, differential abundance analysis was conducted on the training data. Genes were filtered based on specific criteria (adjusted p-value < 0.01, base mean abundance > 100, gene transcript AUC > 0.65, and absolute log2 fold change > 0.25). Filtering criteria were chosen with the intention of selecting <150 genes. Multiple thresholds were tested using our training and validation sample sets and found that the model results were not sensitive to the thresholds chosen. The test set was not used for determining thresholds for gene filtering.

[0097] Raw counts for the training, validation, and test sets were individually normalized using variance stabilizing transformation, and subsets were created based on the chosen transcript features. Variance stabilization transformation was performed using the DESeq2 package and the dispersion function from the training set was used to transform the test and validation sets. Fourteen machine learning classification algorithms were employed using the R package Caret (10.18637 / jss.v028.i05), including generalized linear models with Ridge and LASSO feature selection (GLMNETRIDGE and GLMNETLASSO), support vector machines with linear and radial basis function kernel (SVMLin and SVMRAD), random forest (RF), random forest ExtraTrees (EXTRATREES), neural networks (NNET), linear discriminant analysis (LDA), nearest shrunken centroids (PAM), C5.0 (C5), k-nearest neighbors (KNN), naive bayes (NB), CART (RPART), and generalized linear models (GLM). Training was performed using 5-fold cross-validation and grid search hyperparameter tuning.

[0098] For each model, classification score thresholds were determined using Youden’s index on the training data. The trained models were then employed to predict labels for the validation set, and performance was assessed using the area under the receiver operating characteristic curve area under the curve (ROC-AUC). The model achieving the highest AUC on the validation set was selected as the final model and subsequently applied to the test set, which had not been utilized in any phase of model training or selection. Prediction on both the validation and test sets utilized the Youden’s index threshold derived from the training set.

[0099] Machine learning: Multi-Classification. Samples were partitioned into train and test sets at a ratio of 70:30, considering factors such as diagnosis, severity / sub group, and hospital of origin. Raw counts for the training and test sets were individually normalized using variance stabilizing transformation, as implemented using the DESeq2 package, and the dispersion function from the training set was used to transform the test set. One-vs-one GLMNET LASSO models were trained for each pairwise comparison of samples groups (eg. KD vs Viral Infection) using the top 150 significant features (adjusted p-value <0.05, base mean abundance > 50, and absolute log2 fold change > 1) ordered by gene transcript AUC calculated using the training data. Trained models were used to calculate classification scores for all samples in the train and test data set, resulting in six classification scores for each sample. Classification scores were used to train a multiclass Random Forest algorithm which assigned probability scores for each condition.The final condition with the highest probability score from the Random Forest was assigned as the prediction for each sample.

[0100] CTO analysis. Cell type deconvolution was performed using BayesPrism (vl.l)(48) with the Tabula Sapiens single-cell RNA-seq atlas (Release 1) as a reference. Cells from the Tabula Sapiens atlas were grouped as previously described in Vorperian et al.. Cell types with more than 100,000 unique molecular identifiers (UMIs) were included in the reference and subsampled to 300 cells using ScanPy (vl.8.1). Deconvolution values were scaled from 0-1 for each cell type and medians calculated for plotting.

[0101] For liver damage assessment, samples were separated as having normal or high ALT measurements (normal: ALT < 40 IU / L, high: ALT > 100 IU / L). The ALT measurements were taken from the same blood draw as the plasma for cfRNA processing. For cardiac function, patients were categorized as abnormal if they had abnormal EKG / ECG and / or echocardiogram results. EKG / ECG and echocardiogram results were categorized as abnormal in the context of the patient narrative and final interpretation of the studies. For endothelial damage, samples were separated as either having KD / MIS-C or bacterial / viral infection. CO VID-19 patients were determined to have moderate or severe disease using the following criteria:

[0102] Moderate: The patient must have been hospitalized due to COVID-19 respiratory disease and / or any systemic / non-respiratory symptoms attributed to COVID- 19 (e.g., neonatal fever, dehydration, new diagnosis diabetes, acute appendicitis, necrosis of extremities, diarrhea, encephalopathy, renal insufficiency, mild coagulation abnormalities, etc ).

[0103] Severe: The patient must have been hospitalized for COVID-19 with either high-flow oxygen requirement (high-flow nasal cannula (NC), continuous positive airway pressure (CPAP), bilevel positive airway pressure (BIPAP), intubation with mechanical ventilation, or extracorporeal membranous oxygenation (ECMO)) and / or evidence of end-organ failure (acute renal failure requiring dialysis, coagulation abnormalities resulting in bleeding or stroke, diabetic ketoacidosis (DKA), hemodynamic instability requiring vasopressors) and / or dying from COVID-19. These patients were almost always admitted to the ICU.

[0104] Machine learning: COVID-19 vs non-COVID19 viral infection. Training was done using the same method as the KD vs MIS-C model. Briefly, viral infection samples were partitioned into train and test sets at a ratio of 70:30, considering factors such as COVID-19status, severity / subgroup, and hospital of origin. Subsequently, differential abundance analysis was conducted on the training data. Genes were filtered based on specific criteria (adjusted p- value < 0.01, base mean abundance > 100, gene transcript AUC > 0.65, and absolute log2 fold change > 0.25) and the top 150 genes, as ordered by gene transcript AUC, were selected as inputs for machine learning analysis.

[0105] Raw counts for the train and test sets were individually normalized using variance stabilizing transformation, and subsets were created based on the chosen transcript features. A GLMNET with LASSO regression was trained on the training set using 5-fold cross-validation and grid search hyperparameter tuning. Classification score thresholds were determined using Youden’s index on the training data. The trained models were then employed to predict labels for the test set using the Youden’s index threshold derived from the training set.

[0106] Quantification and statistical analyses. The programming language R (v4.1.0) was utilized for all statistical analyses. Statistical significance was assessed through two-sided Wilcoxon signed-rank tests and Mann- Whitney U tests, unless specified otherwise. Machine learning algorithms were trained using the Caret R package and pipelines were run using the Snakemake workflow management system. In boxplots, boxes denote the 25th and 75th percentiles, the band within the box signifies the median, and whiskers extend to 1.5 times the interquartile range of the hinge. The alignment of all sequencing data was performed against the GRCh38 Gencode v38 Primary Assembly, with feature counting conducted using the GRCh38 Gencode v38 Primary Assembly Annotation.

[0107] Data availability. Raw sequencing data in this study cannot be deposited in a public repository due to patient privacy concerns and lack of consent for a subset of the patient samples. Instead, de-identified RNA-seq count matrices have been uploaded to the NCBI (National Center for Biotechnology Information) GEO (Gene Expression Omnibus) database and will be publicly available upon publication (GSE255555). All code has been deposited on GitHub and will be available upon publication.

Claims

WHAT IS CLAIMED IS:

1. A method comprising: a) obtaining cell-free RNA from a biological sample of a subject, wherein the biological sample is a plasma sample or a serum sample; b) determining a profile of a panel of biomarkers in the cell-free RNA; and c) determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers.

2. The method of claim 1, wherein the step of determining a profile of a panel of biomarkers comprises performing nucleotide sequencing of the cell-free RNA.

3. The method of claim 1 or 2, wherein the step of determining a profile of a panel of biomarkers comprises measuring the levels of RNAs corresponding to the biomarkers in the panel.

4. The method of claim 3, wherein the panel of biomarkers comprises one, more, or all (including, e.g., at least 30, at least 50, at least 75, or at least 100) of genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CT SB, BAZ1A, AHNAK, ANXA6, VCAN, PGD, PHACTR2, S100A8, PFN1, VIM, MIB1, TCF7, NRIP1, CSF3R, MX1, BIRC2, PTGES3, POU2F2, FKBP5, HDLBP, H2BC9, DARS1, AKNA, LCP1, SH3BP5, CCDC88C, DST, EIF4B, MVP, GRB2, HIP1, CD74, NFE2L1, WIPF1, PEAK1, CSK, TBC1D14, PTMA, KDM6B, LYZ, NUCKS1, RAPGEF1, MT-TL1, HIF1A, TXNIP, YBX1, GPI, PKP4, RPL21, RABEP1, TDP2, MAP4K4, GAB ARAP, HELZ, CGNL1, RDX, DENND5A, LY6E, RASSF2, SLA, FXYD5, USF3, TRIM28, IFITM1, SVIL, TRIOBP, MT-TI, RPS5, PRPF8, EEF2, CD164, FLOT1, HSPA8, HBA2, LTA4H, PNRC1, SHANK3, CD48, TNIP1, SBNO2, PHC2, ARRB2, MCTP2, and BCL2Al .

5. The method of claim 4, wherein the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL,SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK, ANXA6, VCAN, PGD, PHACTR2, S100A8, PFN1, VIM, MIB1, TCF7, NRIP1, CSF3R, MX1, BIRC2, PTGES3, POU2F2, FKBP5, HDLBP, H2BC9, DARS1, AKNA, LCP1, SH3BP5, CCDC88C, DST, EIF4B, MVP, GRB2, HIP1, CD74, NFE2L1, WIPF1, PEAK1, CSK, TBC1D14, PTMA, KDM6B, LYZ, NUCKS1, RAPGEF1, MT-TL1, HIF1A, TXNIP, YBX1, GPI, PKP4, RPL21, RABEP1, TDP2, MAP4K4, GABARAP, HELZ, CGNL1, RDX, DENND5A, LY6E, and RASSF2.

6. The method of claim 4, wherein the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNFP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK, ANXA6, VCAN, PGD, PHACTR2, S100A8, PFN1, VIM, MIB1, TCF7, NRIP1, CSF3R, MX1, BIRC2, PTGES3, POU2F2, FKBP5, HDLBP, H2BC9, DARS1, AKNA, LCP1, SH3BP5, CCDC88C, DST, EIF4B, MVP, GRB2, HIP1, and CD74.

7. The method of claim 4, wherein the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNLP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, HEMGN, HECW2, RBM33, S100A4, BCL9L, EIF2S2, CTSB, BAZ1A, AHNAK, ANXA6, VCAN, PGD, PHACTR2, and S100A8.

8. The method of claim 4, wherein the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, CD44, UBA52, SLFN5, GLUL, TRIR, CAI, FTL, SNX3, SAMD9, ATP5F1E, and HEMGN.

9. The method of claim 4, wherein the panel of biomarkers comprises one, more, or all of the genes selected from the group consisting of: AKAP12, DNMT1, IFI27, SUB1, S100A12, PABPC4, ANXA11, BNIP3L, HIPK2, and CD44.

10. The method of any one of claims 4-9, wherein the genes in the panel of biomarkers differentiate between Kawasaki disease (KD), Multisystem Inflammatory Syndrome in Children (MIS-C), viral infections, and bacterial infections.

11. The method of claim 3, wherein the panel of biomarkers comprises one or more genes selected from histone protein coding genes and immune related genes.

12. The method of claim 11, wherein the histone protein coding genes comprise one or more genes selected from the group consisting of Hl-10, Hl-2, Hl-3, Hl-4, Hl-5, H2AC11, H2AC12, H2AC13, H2AC14, H2AC16, H2AC17, H2AC20, H2AC21, H2AC4, H2AC6, H2AX, H2AZ1, H2AZ2, H2BC10, H2BC13, H2BC14, H2BC17, H2BC18, H2BC3, H2BU1, H3-3B, H3C1, H3C10, H3C11, H3C12, H3C2, H3C3, H3C7, H3C8, H4C1, H4C13, H4C2, H4C3, H4C4, H4C5, H4C6, and H4C9.

13. The method of claim 11, wherein the immune related genes comprise neutrophil markers.

14. The method of claim 13, wherein the neutrophil markers comprise one or more genes selected from the group consisting of ABCA13, ACTR10, ADAM10, ALAD, ANO6, ANXA2, ARG1, ARL8A, ARMC8, ATP8A1, AZU1, B2M, BIN2, BPI, BRI3, C5AR1, CAMP, CANT1, CAT, CD177, CD36, CD47, CD53, CD59, CDA, CEACAM1 , CEACAM3, CEACAM6, CEACAM8, CKAP4, CLEC12A, CLEC4D, CLEC5A, CNN2, CPNE3, CTSG, CXCR1, CXCR2, CYBA, CYSTM1, DEFA1, DEFA4, DIAPH1, DNAJC13, DNAJC5, DYNLL1, ELANE, FABP5, FCAR, FCGR3A / FCGR3B, FPR1, FPR2, FTH1, GCA, GDI2, GMFG, GNS, GSTP1, GYG1, HEBP2, HMGB1, HP, HSP90AA1, IDH1, ILF2, IQGAP1, IQGAP2, ITGB2, LAMTOR2, LGALS3, LILRB2, LILRB3, LRG1, LTA4H, LTF, LYZ, MGAM, MGST1, MMP8, MMP9, MNDA, MPO, MS4A3, NME2, PA2G4, PADI2, PGLYRP1, PLAC8, PLAUR, PLD1, PPBP, PRDX6, PRTN3, PSMB7, PSMC2, PSMD3, PSMD7, PTPRJ, PYCARD, PYGB, QPCT, RAB10, RAB14, RAB27A, RAB31, RAB37, RAB3D, RAB6A, RAP1B, RAP2B, RETN, RHOG, RNASE2, S100A11, S100A12, S100A8, S100A9, SIOOP, SELL, SERPINA1, SERPINB10, SIRPA, SIRPB1, SLC2A5, SLC44A2, SLCO4C1, SLPI, SNAP23, SRP14, STOM, SVIP, TCN1, TIMP2, TMBIM1, TMEM63A, TOMI, TRAPPCI, TUBB, TYROBP, VCL, XRCC5, XRCC6, YPEL5, AKT3, ATP5F1B, ATP5PB, C1QB, CASP1, CASP4, CCL5, CCR1,CHP1, C0L17A1, COL24A1, GP1BA, IFNAR1, IGHA1, IGHA2, IGHG1, IGKC, IGLC2, IL1B, ITGA2, ITGB1, ITPR1, ITPR2, JCHAIN, MCL1, MT-ATP6, MT-CYB, MT-ND1, MT- ND2, MT-ND3, MT-ND4, MT-ND4L, MT-ND5, MT-ND6, NCF2, NDUFA5, NDUFA6, NDUFB1, NDUFB2, NDUFB6, NDUFB9, NDUFS6, NLRP3, 0RAI1, PF4, PIK3C3, PIK3CB, PLA2G12A, PLA2G4A, PLAAT1, PLCB4, PLCHI, PNPLA8, PPP3R1, RAC2, SDHB, SELPLG, TIMM 13, TLR4, TSPO, UQCRFS1, VDAC1, VDAC3, VSTM1, ARPC2, ARPC3, CALM1, GNAI3, GNA01, GNAQ, GNAS, GNAZ, GNB1, GNB2, GNB5, GNG10, GNG11, GNG2, GNG5, KRAS, NFAT5, NFKBIA, NFKBIB, RALB, and RELB.

15. The method of claim 3, wherein the step of determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers comprises differentiating between Kawasaki disease (KD) and Multisystem Inflammatory Syndrome in Children (MIS-C) and wherein the panel of biomarkers comprises TXNIP, ASH1L, YPEL5, PCBP1, STK17B, TRAK2, PTMA, RNA5SP149, AFF1, FHDC1, ANXA6, FKBP5, HSP90AB1, SYNE1, RBM33, VIM, PDCD4, CD44, RESF1, DAZAP2, COTL1, PFN1, EEF2, TRIR, and FTL.

16. The method of claim 3, wherein the step of determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers comprises differentiating between Kawasaki disease (KD) and Multisystem Inflammatory Syndrome in Children (MIS-C) and wherein the panel of biomarkers comprises EEF2 and / or FKBP5.

17. The method of claim 16, wherein the panel of biomarkers comprises EEF2 and / or FKBP5 along with one or more additional genes.

18. The method of claim 16, wherein the panel of biomarkers comprises EEF2 and / or FKBP5 along with 2-8 additional genes.

19. The method of any one of the previous claims, wherein determining an inflammatory disease status in the subject based on the profile of the panel of biomarkers comprises comparing the profile of the panel of biomarkers in the cell free RNA from the biological sample to a reference profile of the panel of biomarkers.

20. The method of any one of the previous claims, wherein determining an inflammatory syndrome status in the subject based on the profile of the panel of biomarkers comprises comparing the levels of RNAs corresponding to the biomarkers in the panel to levels of RNAs corresponding to the biomarkers in the reference profile.

21. The method of any one of the previous claims, wherein the inflammatory syndrome comprises Kawasaki disease (KD), Multisystem Inflammatory Syndrome in Children (MIS-C), viral infections, and bacterial infections.

22. The method of any one of the previous claims, wherein determining the inflammatory syndrome status in the subject based on the profile of the panel of biomarkers comprises determining that the subject is more likely to have one inflammatory syndrome as compared to other inflammatory syndromes, e.g., the subject is more likely to have KD as compared to MIS- C, viral infections or bacterial infections.

23. The method of any one of the previous claims, further comprising: d) determining the cell type of origins from cell-free RNA; e) determining tissue damage based on the cell type of origin of the cell-free RNA; and f) determining a clinical decision support tool using the determination of inflammatory disease from the panel of biomarkers and determination of tissue damage from the cell type of origin measurements.

24. The method of claim 23 wherein the step of determining the cell type of origin of cfRNA comprises performing nucleotide sequencing of the cell-free RNA.

25. The method of claim 23 or 24, wherein the step of estimating the cfRNA cell types comprises using a reference RNA-seq data set with a deconvolution algorithm.26 The method of claim 23, 24, or 25, wherein tissue damage is measured by comparing the cell type of origin measurements to a reference group.

27. The method of any one of claims 23-26, wherein the clinical decision support tool comprises the predication of inflammatory condition and measurements of tissue damage.

28. The method of any one of the previous claims, wherein the biological sample is a plasma sample.

29. The method of any one of the previous claims, wherein the biological sample is a serum sample.

30. The method of any one of the previous claims, wherein the subject is a mammal.

31. The method of any one of the previous claims, wherein the subject is a human.

32. The method of any one of the previous claims, wherein the subject is in a pediatric subpopulation of human.

33. The method of any one of the previous claims, wherein the profile of the panel of biomarkers comprises subject-specific gene expression of the biomarkers in the panel.

34. The method of any one of the previous claims, wherein the profile of the panel of biomarkers comprises a score determined based on the levels of RNAs corresponding to the biomarkers in the panel using a trained classification model, and wherein the score correlates with an inflammatory syndrome status.

35. The method of any one of the previous claims, wherein the cell-free RNA is substantially free of contaminant RNA.

36. The method of any one of the previous claims, wherein the inflammatory syndrome status determined is used in determining treatment for the subject.

37. The method of any one of the previous claims, wherein test performance of the trained classification model is measured through one or more of accuracy, sensitivity, specificity, and / or area under the receiver operating characteristic curve (ROC-AUC).

38. The method of any one of the previous claims, wherein the inflammatory syndrome status is determined at a greater than 80%, 85%, 90%, 91%, 92%, 93%, 94%, or 95% accuracy.

39. The method of any one of the previous claims, wherein the inflammatory syndrome status is determined at a greater than 80%, 85%, 90%, 91%, 92%, 93%, 94%, or 95% sensitivity.

40. The method of any one of the previous claims, wherein the inflammatory syndrome status is determined at a greater than 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, or 95% specificity.

41. The method of any one of the previous claims, wherein the trained classification model comprises one or more models selected from the group consisting of generalized linear models with Ridge and LASSO feature selection (GLMNETRIDGE and GLMNETLASSO), support vector machines with linear and radial basis function kernel (SVMLin and SVMRAD), random forest (RF), random forest ExtraTrees (EXTRATREES), neural networks (NNET), linear discriminant analysis (LDA), nearest shrunken centroids (PAM), C5.0 (C5), k-nearest neighbors (KNN), naive bayes (NB), CART (RPART), generalized linear model (GLM), and greedy forward search algorithm (GFS).

42. The method of any one of the previous claims, wherein the method further comprises a step of treating the subject based on the inflammatory syndrome status.

43. The method of claim 42, wherein the inflammatory syndrome status is determined to be Kawasaki disease (KD), the step of treating the subject based on the inflammatory syndrome comprises treating the subject with intravenous immunoglobulin (IVIG).

44. The method of claim 42, wherein the inflammatory syndrome status is determined to be Multisystem Inflammatory Syndrome in Children (MIS-C), the step of treating the subject based on the inflammatory syndrome comprises treating the subject with IVIG and a low to moderate dose glucocorticoid.

45. The method of claim 42, wherein the inflammatory syndrome status is determined to be a bacterial infection, the step of treating the subject based on the inflammatory syndrome comprises treating the subject with an antibiotic.

46. The method of claim 42, wherein the inflammatory syndrome status is determined to be a viral infection, the step of treating the subject based on the inflammatory syndrome comprises treating the subject with a normal treatment plan for such a virus.

47. The method of any one previous claim wherein the subject is immunocompromised.

Images