An interactive atlas of genomic, proteomic, and metabolomic biomarkers promotes the potential of proteins to predict complex diseases.

Multiomics analyses have identified multiple potential biomarkers of the incidence and prevalence of complex diseases. However, it is not known which type of biomarker is optimal for clinical purposes. Here, we make a systematic comparison of 90 million genetic variants, 1453 proteins, and 325 metabolites from 500,000 individuals with complex diseases from the UK Biobank. A machine learning pipeline consisting of data cleaning, data imputation, feature selection, and model training using cross-validation and comparison of the results on holdout test sets showed that proteins were most predictive, followed by metabolites, and genetic variants. Only five proteins per disease resulted in median (min-max) areas under the receiver operating characteristic curves for incidence of 0.79 (0.65-0.86) and 0.84 (0.70-0.91) for prevalence. In summary, our work suggests the potential of predicting complex diseases based on a limited number of proteins. We provide an interactive atlas (macd.shinyapps.io/ShinyApp/) to find genomic, proteomic, or metabolomic biomarkers for different complex diseases.

Transcript and protein signatures derived from shared molecular interactions across cancers are associated with mortality.

BACKGROUND: Characterization of shared cancer mechanisms have been proposed to improve therapy strategies and prognosis. Here, we aimed to identify shared cell-cell interactions (CCIs) within the tumor microenvironment across multiple solid cancers and assess their association with cancer mortality. METHODS: CCIs of each cancer were identified by NicheNet analysis of single-cell RNA sequencing data from breast, colon, liver, lung, and ovarian cancers. These CCIs were used to construct a shared multi-cellular tumor model (shared-MCTM) representing common CCIs across cancers. A gene signature was identified from the shared-MCTM and tested on the mRNA and protein level in two large independent cohorts: The Cancer Genome Atlas (TCGA, 9185 tumor samples and 727 controls across 22 cancers) and UK biobank (UKBB, 10,384 cancer patients and 5063 controls with proteomics data across 17 cancers). Cox proportional hazards models were used to evaluate the association of the signature with 10-year all-cause mortality, including sex-specific analysis. RESULTS: A shared-MCTM was derived from five individual cancers. A shared gene signature was extracted from this shared-MCTM and the most prominent regulatory cell type, matrix cancer-associated fibroblast (mCAF). The signature exhibited significant expression changes in multiple cancers compared to controls at both mRNA and protein levels in two independent cohorts. Importantly, it was significantly associated with mortality in cancer patients in both cohorts. The highest hazard ratios were observed for brain cancer in TCGA (HR [95%CI] = 6.90[4.64-10.25]) and ovarian cancer in UKBB (5.53[2.08-8.80]). Sex-specific analysis revealed distinct risks, with a higher mortality risk associated with the protein signature score in males (2.41[1.97-2.96]) compared to females (1.84[1.44-2.37]). CONCLUSION: We identified a gene signature from a comprehensive shared-MCTM representing common CCIs across different cancers and revealed the regulatory role of mCAF in the tumor microenvironment. The pathogenic relevance of the gene signature was supported by differential expression and association with mortality on both mRNA and protein levels in two independent cohorts.

scDrugPrio: a framework for the analysis of single-cell transcriptomics to address multiple problems in precision medicine in immune-mediated inflammatory diseases.

BACKGROUND: Ineffective drug treatment is a major problem for many patients with immune-mediated inflammatory diseases (IMIDs). Important reasons are the lack of systematic solutions for drug prioritisation and repurposing based on characterisation of the complex and heterogeneous cellular and molecular changes in IMIDs. METHODS: Here, we propose a computational framework, scDrugPrio, which constructs network models of inflammatory disease based on single-cell RNA sequencing (scRNA-seq) data. scDrugPrio constructs detailed network models of inflammatory diseases that integrate information on cell type-specific expression changes, altered cellular crosstalk and pharmacological properties for the selection and ranking of thousands of drugs. RESULTS: scDrugPrio was developed using a mouse model of antigen-induced arthritis and validated by improved precision/recall for approved drugs, as well as extensive in vitro, in vivo, and in silico studies of drugs that were predicted, but not approved, for the studied diseases. Next, scDrugPrio was applied to multiple sclerosis, Crohn's disease, and psoriatic arthritis, further supporting scDrugPrio through prioritisation of relevant and approved drugs. However, in contrast to the mouse model of arthritis, great interindividual cellular and gene expression differences were found in patients with the same diagnosis. Such differences could explain why some patients did or did not respond to treatment. This explanation was supported by the application of scDrugPrio to scRNA-seq data from eleven individual Crohn's disease patients. The analysis showed great variations in drug predictions between patients, for example, assigning a high rank to anti-TNF treatment in a responder and a low rank in a nonresponder to that treatment. CONCLUSIONS: We propose a computational framework, scDrugPrio, for drug prioritisation based on scRNA-seq of IMID disease. Application to individual patients indicates scDrugPrio's potential for personalised network-based drug screening on cellulome-, genome-, and drugome-wide scales. For this purpose, we made scDrugPrio into an easy-to-use R package ( https://github.com/SDTC-CPMed/scDrugPrio ).

An interactive atlas of genomic, proteomic, and metabolomic biomarkers promotes the potential of proteins to predict complex diseases.

Multiomics analyses have identified multiple potential biomarkers of the incidence and prevalence of complex diseases. However, it is not known which type of biomarker is optimal for clinical purposes. Here, we make a systematic comparison of 90 million genetic variants, 1,453 proteins, and 325 metabolites from 500,000 individuals with complex diseases from the UK Biobank. A machine learning pipeline consisting of data cleaning, data imputation, feature selection, and model training using cross-validation and comparison of the results on holdout test sets showed that proteins were most predictive, followed by metabolites, and genetic variants. Only five proteins per disease resulted in median (min-max) areas under the receiver operating characteristic curves for incidence of 0.79 (0.65-0.86) and 0.84 (0.70-0.91) for prevalence. In summary, our work suggests the potential of predicting complex diseases based on a limited number of proteins. We provide an interactive atlas (macd.shinyapps.io/ShinyApp/) to find genomic, proteomic, or metabolomic biomarkers for different complex diseases.

scDrugPrio: A framework for the analysis of single-cell transcriptomics to address multiple problems in precision medicine in immune-mediated inflammatory diseases.

Background: Ineffective drug treatment is a major problem for many patients with immune-mediated inflammatory diseases (IMIDs). Important reasons are the lack of systematic solutions for drug prioritisation and repurposing based on characterisation of the complex and heterogeneous cellular and molecular changes in IMIDs. Methods: Here, we propose a computational framework, scDrugPrio, which constructs network models of inflammatory disease based on single-cell RNA sequencing (scRNA-seq) data. scDrugPrio constructs detailed network models of inflammatory diseases that integrate information on cell type-specific expression changes, altered cellular crosstalk and pharmacological properties for the selection and ranking of thousands of drugs. Results: scDrugPrio was developed using a mouse model of antigen-induced arthritis and validated by improved precision/recall for approved drugs, as well as extensive in vitro, in vivo, and in silico studies of drugs that were predicted, but not approved, for the studied diseases. Next, scDrugPrio was applied to multiple sclerosis, Crohn's disease, and psoriatic arthritis, further supporting scDrugPrio through prioritisation of relevant and approved drugs. However, in contrast to the mouse model of arthritis, great interindividual cellular and gene expression differences were found in patients with the same diagnosis. Such differences could explain why some patients did or did not respond to treatment. This explanation was supported by the application of scDrugPrio to scRNA-seq data from eleven individual Crohn's disease patients. The analysis showed great variations in drug predictions between patients, for example, assigning a high rank to anti-TNF treatment in a responder and a low rank in a nonresponder to that treatment. Conclusion: We propose a computational framework, scDrugPrio, for drug prioritisation based on scRNA-seq of IMID disease. Application to individual patients indicates scDrugPrio's potential for personalised network-based drug screening on cellulome-, genome-, and drugome-wide scales. For this purpose, we made scDrugPrio into an easy-to-use R package (https://github.com/SDTC-CPMed/scDrugPrio).

Multi-organ single-cell analysis reveals an on/off switch system with potential for personalized treatment of immunological diseases.

Prioritization of disease mechanisms, biomarkers, and drug targets in immune-mediated inflammatory diseases (IMIDs) is complicated by altered interactions between thousands of genes. Our multi-organ single-cell RNA sequencing of a mouse IMID model, namely collagen-induced arthritis, shows highly complex and heterogeneous expression changes in all analyzed organs, even though only joints showed signs of inflammation. We organized those into a multi-organ multicellular disease model, which shows predicted molecular interactions within and between organs. That model supports that inflammation is switched on or off by altered balance between pro- and anti-inflammatory upstream regulators (URs) and downstream pathways. Meta-analyses of human IMIDs show a similar, but graded, on/off switch system. This system has the potential to prioritize, diagnose, and treat optimal combinations of URs on the levels of IMIDs, subgroups, and individual patients. That potential is supported by UR analyses in more than 600 sera from patients with systemic lupus erythematosus.

Digital Twins for Predictive, Preventive Personalized, and Participatory Treatment of Immune-Mediated Diseases.

Digital twins are computational models of complex systems, which aim to understand and optimize those systems more effectively than would be possible in real life. Ideally, digital twins can be translated to individual patients, to characterize and computationally treat their diseases with thousands of drugs, to select the drug or drugs that cure the patients. The background problem is that many patients do not respond adequately to drug treatment. This problem reflects a wide gap between the complexity of diseases and clinical practice. Each disease may involve altered interactions between thousands of genes that vary between different cell types in different organs. To our knowledge, these altered interactions have not been characterized on a genome-, cellulome-, and organ-wide scale in any disease. Thus, clinical translation of the digital twin ideal for predictive, preventive, personalized and participatory treatment involves formidable challenges, which are close to the limits of, or beyond today's technologies. Here, I discuss recent developments and challenges in relation to that ideal focusing on immune-mediated inflammatory diseases, as well as examples from other diseases.

RNA-sequencing and mass-spectrometry proteomic time-series analysis of T-cell differentiation identified multiple splice variants models that predicted validated protein biomarkers in inflammatory diseases.

Profiling of mRNA expression is an important method to identify biomarkers but complicated by limited correlations between mRNA expression and protein abundance. We hypothesised that these correlations could be improved by mathematical models based on measuring splice variants and time delay in protein translation. We characterised time-series of primary human naïve CD4+ T cells during early T helper type 1 differentiation with RNA-sequencing and mass-spectrometry proteomics. We performed computational time-series analysis in this system and in two other key human and murine immune cell types. Linear mathematical mixed time delayed splice variant models were used to predict protein abundances, and the models were validated using out-of-sample predictions. Lastly, we re-analysed RNA-seq datasets to evaluate biomarker discovery in five T-cell associated diseases, further validating the findings for multiple sclerosis (MS) and asthma. The new models significantly out-performing models not including the usage of multiple splice variants and time delays, as shown in cross-validation tests. Our mathematical models provided more differentially expressed proteins between patients and controls in all five diseases. Moreover, analysis of these proteins in asthma and MS supported their relevance. One marker, sCD27, was validated in MS using two independent cohorts for evaluating response to treatment and disease prognosis. In summary, our splice variant and time delay models substantially improved the prediction of protein abundance from mRNA expression in three different immune cell types. The models provided valuable biomarker candidates, which were further validated in MS and asthma.

A dynamic single cell-based framework for digital twins to prioritize disease genes and drug targets.

BACKGROUND: Medical digital twins are computational disease models for drug discovery and treatment. Unresolved problems include how to organize and prioritize between disease-associated changes in digital twins, on cellulome- and genome-wide scales. We present a dynamic framework that can be used to model such changes and thereby prioritize upstream regulators (URs) for biomarker- and drug discovery. METHODS: We started with seasonal allergic rhinitis (SAR) as a disease model, by analyses of in vitro allergen-stimulated peripheral blood mononuclear cells (PBMC) from SAR patients. Time-series a single-cell RNA-sequencing (scRNA-seq) data of these cells were used to construct multicellular network models (MNMs) at each time point of molecular interactions between cell types. We hypothesized that predicted molecular interactions between cell types in the MNMs could be traced to find an UR gene, at an early time point. We performed bioinformatic and functional studies of the MNMs to develop a scalable framework to prioritize UR genes. This framework was tested on a single-cell and bulk-profiling data from SAR and other inflammatory diseases. RESULTS: Our scRNA-seq-based time-series MNMs of SAR showed thousands of differentially expressed genes (DEGs) across multiple cell types, which varied between time points. Instead of a single-UR gene in each MNM, we found multiple URs dispersed across the cell types. Thus, at each time point, the MNMs formed multi-directional networks. The absence of linear hierarchies and time-dependent variations in MNMs complicated the prioritization of URs. For example, the expression and functions of Th2 cytokines, which are approved drug targets in allergies, varied across cell types, and time points. Our analyses of bulk- and single-cell data from other inflammatory diseases also revealed multi-directional networks that showed stage-dependent variations. We therefore developed a quantitative approach to prioritize URs: we ranked the URs based on their predicted effects on downstream target cells. Experimental and bioinformatic analyses supported that this kind of ranking is a tractable approach for prioritizing URs. CONCLUSIONS: We present a scalable framework for modeling dynamic changes in digital twins, on cellulome- and genome-wide scales, to prioritize UR genes for biomarker and drug discovery.

CD4+ T-cell DNA methylation changes during pregnancy significantly correlate with disease-associated methylation changes in autoimmune diseases.

Epigenetics may play a central, yet unexplored, role in the profound changes that the maternal immune system undergoes during pregnancy and could be involved in the pregnancy-induced modulation of several autoimmune diseases. We investigated changes in the methylome in isolated circulating CD4+ T-cells in non-pregnant and pregnant women, during the 1st and 2nd trimester, using the Illumina Infinium Human Methylation 450K array, and explored how these changes were related to autoimmune diseases that are known to be affected during pregnancy. Pregnancy was associated with several hundreds of methylation differences, particularly during the 2nd trimester. A network-based modular approach identified several genes, e.g., CD28, FYN, VAV1 and pathways related to T-cell signalling and activation, highlighting T-cell regulation as a central component of the observed methylation alterations. The identified pregnancy module was significantly enriched for disease-associated methylation changes related to multiple sclerosis, rheumatoid arthritis and systemic lupus erythematosus. A negative correlation between pregnancy-associated methylation changes and disease-associated changes was found for multiple sclerosis and rheumatoid arthritis, diseases that are known to improve during pregnancy whereas a positive correlation was found for systemic lupus erythematosus, a disease that instead worsens during pregnancy. Thus, the directionality of the observed changes is in line with the previously observed effect of pregnancy on disease activity. Our systems medicine approach supports the importance of the methylome in immune regulation of T-cells during pregnancy. Our findings highlight the relevance of using pregnancy as a model for understanding and identifying disease-related mechanisms involved in the modulation of autoimmune diseases.Abbreviations: BMIQ: beta-mixture quantile dilation; DMGs: differentially methylated genes; DMPs: differentially methylated probes; FE: fold enrichment; FDR: false discovery rate; GO: gene ontology; GWAS: genome-wide association studies; MDS: multidimensional scaling; MS: multiple sclerosis; PBMC: peripheral blood mononuclear cells; PBS: phosphate buffered saline; PPI; protein-protein interaction; RA: rheumatoid arthritis; SD: standard deviation; SLE: systemic lupus erythematosus; SNP: single nucleotide polymorphism; TH: CD4+ T helper cell; VIStA: diVIsive Shuffling Approach.

TET2 as a tumor suppressor and therapeutic target in T-cell acute lymphoblastic leukemia.

Pediatric T-cell acute lymphoblastic leukemia (T-ALL) is an aggressive malignancy resulting from overproduction of immature T-cells in the thymus and is typified by widespread alterations in DNA methylation. As survival rates for relapsed T-ALL remain dismal (10 to 25%), development of targeted therapies to prevent relapse is key to improving prognosis. Whereas mutations in the DNA demethylating enzyme TET2 are frequent in adult T-cell malignancies, TET2 mutations in T-ALL are rare. Here, we analyzed RNA-sequencing data of 321 primary T-ALLs, 20 T-ALL cell lines, and 25 normal human tissues, revealing that TET2 is transcriptionally repressed or silenced in 71% and 17% of T-ALL, respectively. Furthermore, we show that TET2 silencing is often associated with hypermethylation of the TET2 promoter in primary T-ALL. Importantly, treatment with the DNA demethylating agent, 5-azacytidine (5-aza), was significantly more toxic to TET2-silenced T-ALL cells and resulted in stable re-expression of the TET2 gene. Additionally, 5-aza led to up-regulation of methylated genes and human endogenous retroviruses (HERVs), which was further enhanced by the addition of physiological levels of vitamin C, a potent enhancer of TET activity. Together, our results clearly identify 5-aza as a potential targeted therapy for TET2-silenced T-ALL.

[Clinical translation of genomic medicine]. / Stor potential när genomikdata kan implementeras i klinisk rutin.

Recent technical developments and early clinical examples support that precision medicine has potential to provide novel diagnostic and therapeutic solutions for patients with complex diseases, who are not responding to existing therapies. Those solutions will require integration of genomic data with routine clinical, imaging, sensor, biobank and registry data. Moreover, user-friendly tools for informed decision support for both patients and clinicians will be needed. While this will entail huge technical, ethical, societal and regulatory challenges, it may contribute to transforming and improving health care towards becoming predictive, preventive, personalised and participatory (4P-medicine).

[Precision medicine in complex diseases]. / Precisionsmedicin kan bli viktig även vid komplexa sjukdomar.

Complex diseases represent a number of common disorders such as allergic conditions, cardiovascular, metabolic and chronic inflammatory diseases. These diseases are caused by a combination of genetic, environmental and lifestyle factors. This complex etiology creates challenges when it comes to diagnostics, follow-up programs and treatment. Although exact disease mechanisms are yet to be elucidated for most complex diseases, key genetic determinants have been mapped and omics profiling has unraveled involved pathways. Using this wealth of data, precision medicine applications have started to appear also for common, complex diseases. In this article, we review current precision medicine applications from a clinical point of view and outline briefly a roadmap ahead for further clinical implementation of precision medicine in complex diseases.

Meta-Analysis of Expression Profiling Data Indicates Need for Combinatorial Biomarkers in Pediatric Ulcerative Colitis.

BACKGROUND: Unbiased studies using different genome-wide methods have identified a great number of candidate biomarkers for diagnosis and treatment response in pediatric ulcerative colitis (UC). However, clinical translation has been proven difficult. Here, we hypothesized that one reason could be differences between inflammatory responses in an inflamed gut and in peripheral blood cells. METHODS: We performed meta-analysis of gene expression microarray data from intestinal biopsies and whole blood cells (WBC) from pediatric patients with UC and healthy controls in order to identify overlapping pathways, predicted upstream regulators, and potential biomarkers. RESULTS: Analyses of profiling datasets from colonic biopsies showed good agreement between different studies regarding pathways and predicted upstream regulators. The most activated predicted upstream regulators included TNF, which is known to have a key pathogenic and therapeutic role in pediatric UC. Despite this, the expression levels of TNF were increased in neither colonic biopsies nor WBC. A potential explanation was increased expression of TNFR2, one of the membrane-bound receptors of TNF in the inflamed colon. Further analyses showed a similar pattern of complex relations between the expression levels of the regulators and their receptors. We also found limited overlap between pathways and predicted upstream regulators in colonic biopsies and WBC. An extended search including all differentially expressed genes that overlapped between colonic biopsies and WBC only resulted in identification of three potential biomarkers involved in the regulation of intestinal inflammation. However, two had been previously proposed in adult inflammatory bowel diseases (IBD), namely, MMP9 and PROK2. CONCLUSIONS: Our findings indicate that biomarker identification in pediatric UC is complicated by the involvement of multiple pathways, each of which includes many different types of genes in the blood or inflamed intestine. Therefore, further studies for identification of combinatorial biomarkers are warranted. Our study may provide candidate biomarkers for such studies.

Correction to: A validated single-cell-based strategy to identify diagnostic and therapeutic targets in complex diseases.

An amendment to this paper has been published and can be accessed via the original article.

Bulk and single cell transcriptomic data indicate that a dichotomy between inflammatory pathways in peripheral blood and arthritic joints complicates biomarker discovery.

BACKGROUND: Unbiased studies using different genome-wide methods have identified several novel biomarkers for diagnosis and treatment response in Rheumatoid Arthritis (RA). However, clinical translation has proven difficult. Here, we hypothesized that one reason could be that inflammatory responses in peripheral blood are different from those in the arthritic joint. METHODS: We performed meta-analysis of gene expression microarray data from synovium, whole blood cells (WBC), peripheral blood mononuclear cells (PBMC), and CD4+ T cells from patients with RA and healthy controls in order to identify overlapping pathways, upstream regulators and potential biomarkers. We also analyzed single cell RNA-sequencing (scRNA-seq) data from peripheral blood and whole joints from a mouse model of antigen-induced arthritis. RESULTS: Analyses of two profiling data sets from synovium from RA patients and healthy controls all showed significant activation of pathways with known pathogenic relevance, such as the Th1 pathway, the role of NFAT in regulation of the immune response, dendritic cell maturation, iCOS-iCOSL signaling in T helper cells, Fcγ receptor-mediated phagocytosis, interferon signaling, Cdc42 signaling, and cytotoxic T lymphocyte-mediated apoptosis. The most activated upstream regulators included TNF, an important drug target, as well as IFN-gamma and CD40LG, all of which are known to play important pathogenic roles in RA. The differentially expressed genes from synovium included several potential biomarkers, such as CCL5, CCL13, CCL18, CX3CL1, CXCL6, CXCL9, CXCL10, CXCL13, IL15, IL32, IL1RN, SPP1, and TNFSF11. By contrast, microarray studies of WBC, PBMC and CD4+ T cells showed variable pathways and limited pathway overlap with synovium. Similarly, scRNA-seq data from a mouse model of arthritis did not support that inflammatory responses in peripheral blood reflect those in the arthritic joints. These data showed pathway overlap between mouse joint cells and synovium from patients with RA, but not with cells in peripheral blood. CONCLUSIONS: Our findings indicate a dichotomy between gene expression changes, pathways, upstream regulators and biomarkers in synovium and cell types in peripheral blood, which complicates identification of biomarkers in blood.

Estimating heritability and genetic correlations from large health datasets in the absence of genetic data.

Typically, estimating genetic parameters, such as disease heritability and between-disease genetic correlations, demands large datasets containing all relevant phenotypic measures and detailed knowledge of family relationships or, alternatively, genotypic and phenotypic data for numerous unrelated individuals. Here, we suggest an alternative, efficient estimation approach through the construction of two disease metrics from large health datasets: temporal disease prevalence curves and low-dimensional disease embeddings. We present eleven thousand heritability estimates corresponding to five study types: twins, traditional family studies, health records-based family studies, single nucleotide polymorphisms, and polygenic risk scores. We also compute over six hundred thousand estimates of genetic, environmental and phenotypic correlations. Furthermore, we find that: (1) disease curve shapes cluster into five general patterns; (2) early-onset diseases tend to have lower prevalence than late-onset diseases (Spearman's ρ = 0.32, p < 10-16); and (3) the disease onset age and heritability are negatively correlated (ρ = -0.46, p < 10-16).

An algorithm-based meta-analysis of genome- and proteome-wide data identifies a combination of potential plasma biomarkers for colorectal cancer.

Screening programs for colorectal cancer (CRC) often rely on detection of blood in stools, which is unspecific and leads to a large number of colonoscopies of healthy subjects. Painstaking research has led to the identification of a large number of different types of biomarkers, few of which are in general clinical use. Here, we searched for highly accurate combinations of biomarkers by meta-analyses of genome- and proteome-wide data from CRC tumors. We focused on secreted proteins identified by the Human Protein Atlas and used our recently described algorithms to find optimal combinations of proteins. We identified nine proteins, three of which had been previously identified as potential biomarkers for CRC, namely CEACAM5, LCN2 and TRIM28. The remaining proteins were PLOD1, MAD1L1, P4HA1, GNS, C12orf10 and P3H1. We analyzed these proteins in plasma from 80 patients with newly diagnosed CRC and 80 healthy controls. A combination of four of these proteins, TRIM28, PLOD1, CEACAM5 and P4HA1, separated a training set consisting of 90% patients and 90% of the controls with high accuracy, which was verified in a test set consisting of the remaining 10%. Further studies are warranted to test our algorithms and proteins for early CRC diagnosis.

A validated single-cell-based strategy to identify diagnostic and therapeutic targets in complex diseases.

BACKGROUND: Genomic medicine has paved the way for identifying biomarkers and therapeutically actionable targets for complex diseases, but is complicated by the involvement of thousands of variably expressed genes across multiple cell types. Single-cell RNA-sequencing study (scRNA-seq) allows the characterization of such complex changes in whole organs. METHODS: The study is based on applying network tools to organize and analyze scRNA-seq data from a mouse model of arthritis and human rheumatoid arthritis, in order to find diagnostic biomarkers and therapeutic targets. Diagnostic validation studies were performed using expression profiling data and potential protein biomarkers from prospective clinical studies of 13 diseases. A candidate drug was examined by a treatment study of a mouse model of arthritis, using phenotypic, immunohistochemical, and cellular analyses as read-outs. RESULTS: We performed the first systematic analysis of pathways, potential biomarkers, and drug targets in scRNA-seq data from a complex disease, starting with inflamed joints and lymph nodes from a mouse model of arthritis. We found the involvement of hundreds of pathways, biomarkers, and drug targets that differed greatly between cell types. Analyses of scRNA-seq and GWAS data from human rheumatoid arthritis (RA) supported a similar dispersion of pathogenic mechanisms in different cell types. Thus, systems-level approaches to prioritize biomarkers and drugs are needed. Here, we present a prioritization strategy that is based on constructing network models of disease-associated cell types and interactions using scRNA-seq data from our mouse model of arthritis, as well as human RA, which we term multicellular disease models (MCDMs). We find that the network centrality of MCDM cell types correlates with the enrichment of genes harboring genetic variants associated with RA and thus could potentially be used to prioritize cell types and genes for diagnostics and therapeutics. We validated this hypothesis in a large-scale study of patients with 13 different autoimmune, allergic, infectious, malignant, endocrine, metabolic, and cardiovascular diseases, as well as a therapeutic study of the mouse arthritis model. CONCLUSIONS: Overall, our results support that our strategy has the potential to help prioritize diagnostic and therapeutic targets in human disease.