Differential Network Analysis Research Articles

Joint Learning of Multiple Differential Networks With Latent Variables.

Graphical models have been widely used to learn the conditional dependence structures among random variables. In many controlled experiments, such as the studies of disease or drug effectiveness, learning the structural changes of graphical models under two different conditions is of great importance. However, most existing graphical models are developed for estimating a single graph and based on a tacit assumption that there is no missing relevant variables, which wastes the common information provided by multiple heterogeneous data sets and underestimates the influence of latent/unobserved relevant variables. In this paper, we propose a joint differential network analysis (JDNA) model to jointly estimate multiple differential networks with latent variables from multiple data sets. The JDNA model is built on a penalized D-trace loss function, with group lasso or generalized fused lasso penalties. We implement a proximal gradient-based alternating direction method of multipliers to tackle the corresponding convex optimization problems. Extensive simulation experiments demonstrate that JDNA model outperforms state-of-the-art methods in estimating the structural changes of graphical models. Moreover, a series of experiments on several real-world data sets have been performed and experiment results consistently show that our proposed JDNA model is effective in identifying differential networks under different conditions.

A novel gene-expression-signature-based model for prediction of response to Tripterysium glycosides tablet for rheumatoid arthritis patients

BackgroundApproximately 30% of rheumatoid arthritis (RA) patients treated with Tripterysium glycosides (TG) tablets fail to achieve clinical improvement, implying the essentiality of predictive biomarkers and tools. Herein, we aimed to identify possible biomarkers predictive of therapeutic effects of TG tablets in RA.MethodsGene expression profile in peripheral blood mononuclear cells obtained from a discovery cohort treated with TG tablets was detected by Affymetrix EG1.0 arrays. Then, a list of candidate gene biomarkers of response to TG tablets were identified by integrating differential expression data analysis and gene signal transduction network analysis. After that, a partial-least-squares (PLS) model based on the expression levels of the candidate gene biomarkers in RA patients was constructed and evaluated using a validation cohort.ResultsSix candidate gene biomarkers (MX1, OASL, SPINK1, CRK, GRAPL and RNF2) were identified to be predictors of TG therapy. Following the construction of a PLS-based model using their expression levels in peripheral blood, both the 5-fold cross-validation and independent dataset validations showed the high predictive efficiency of this model, and demonstrated a distinguished improvement of the PLS-model based on six candidate gene biomarkers’ expression in combination over the commonly used clinical and inflammatory parameters, as well as the gene biomarkers alone, in predicting RA patients’ response to TG tablets.ConclusionsThis hypothesis-generating study identified MX1, OASL, SPINK1, CRK, GRAPL and RNF2 as novel targets for RA therapeutic intervention, and the PLS model based on the expression levels of these candidate biomarkers may have a potential prognostic value in RA patients treated with TG tablets.

Identifying progressive gene network perturbation from single-cell RNA-seq data.

Identifying the gene regulatory networks that control development or disease is one of the most important problems in biology. Here, we introduce a computational approach, called PIPER (ProgressIve network PERturbation), to identify the perturbed genes that drive differences in the gene regulatory network across different points in a biological progression. PIPER employs algorithms tailor-made for single cell RNA sequencing (scRNA-seq) data to jointly identify gene networks for multiple progressive conditions. It then performs differential network analysis along the identified gene networks to identify master regulators. We demonstrate that PIPER outperforms state-of-the-art alternative methods on simulated data and is able to predict known key regulators of differentiation on real scRNA-Seq datasets.

Network-Based Predictors of Progression in Head and Neck Squamous Cell Carcinoma.

The heterogeneity in head and neck squamous cell carcinoma (HNSCC) has made reliable stratification extremely challenging. Behavioral risk factors such as smoking and alcohol consumption contribute to this heterogeneity. To help elucidate potential mechanisms of progression in HNSCC, we focused on elucidating patterns of gene interactions associated with tumor progression. We performed de-novo gene co-expression network inference utilizing 229 patient samples from The Cancer Genome Atlas (TCGA) previously annotated by Bornstein et al. (2016). Differential network analysis allowed us to contrast progressor and non-progressor cohorts. Beyond standard differential expression (DE) analysis, this approach evaluates changes in gene expression variance (differential variability DV) and changes in covariance, which we denote as differential wiring (DW). The set of affected genes was overlaid onto the co-expression network, identifying 12 modules significantly enriched in DE, DV, and/or DW genes. Additionally, we identified modules correlated with behavioral measures such as alcohol consumption and smoking status. In the module enriched for differentially wired genes, we identified network hubs including IL10RA, DOK2, APBB1IP, UBASH3A, SASH3, CELF2, TRAF3IP3, GIMAP6, MYO1F, NCKAP1L, WAS, FERMT3, SLA, SELPLG, TNFRSF1B, WIPF1, AMICA1, PTPN22; the network centrality and progression specificity of these genes suggest a potential role in tumor evolution mechanisms related to inflammation and microenvironment. The identification of this network-based gene signature could be further developed to guide progression stratification, highlighting how network approaches may help improve clinical research end points and ultimately aid in clinical utility.

Differential Co-expression and Regulatory Network Analysis Uncover the Relapse Factor and Mechanism of T Cell Acute Leukemia

The pediatric T cell acute lymphoblastic leukemia (T-ALL) still remains a cancer with worst prognosis for high recurrence. Massive studies were conducted for the leukemia relapse based on diagnosis and relapse paired samples. However, the initially diagnostic samples may contain the relapse information and mechanism, which were rarely studied. In this study, we collected mRNA and microRNA (miRNA) data from initially diagnosed pediatric T-ALL samples with their relapse or remission status after treatment. Integrated differential co-expression and miRNA-transcription factor (TF)-gene regulatory network analyses were used to reveal the possible relapse mechanisms for pediatric T-ALL. We detected miR-1246/1248 and NOTCH2 served as key nodes in the relapse network, and they combined with TF WT1/SOX4/REL to form regulatory modules that influence the progress of T-ALL. A regulatory loop miR-429-MYCN-MFHAS1 was found potentially associated with the remission of T-ALL. Furthermore, we proved miR-1246/1248 combined with NOTCH2 could promote cell proliferation in the T-ALL cell line by experiments. Meanwhile, analysis based on the miRNA-drug relationships demonstrated that drugs 5-fluorouracil, ascorbate, and trastuzumab targeting miR-1246 could serve as potential supplements for the standard therapy. In conclusion, our findings revealed the potential molecular mechanisms of T-ALL relapse by the combination of co-expression and regulatory network, and they provide preliminary clues for precise treatment of T-ALL patients.

Loss of Trem2 in microglia leads to widespread disruption of cell coexpression networks in mouse brain.

Rare heterozygous coding variants in the triggering receptor expressed in myeloid cells 2 (TREM2) gene, conferring increased risk of developing late-onset Alzheimer's disease, have been identified. We examined the transcriptional consequences of the loss of Trem2 in mouse brain to better understand its role in disease using differential expression and coexpression network analysis of Trem2 knockout and wild-type mice. We generated RNA-Seq data from cortex and hippocampus sampled at 4 and 8 months. Using brain cell-type markers and ontology enrichment, we found subnetworks with cell type and/or functional identity. We primarily discovered changes in an endothelial gene-enriched subnetwork at 4 months, including a shift toward a more central role for the amyloid precursor protein gene, coupled with widespread disruption of other cell-type subnetworks, including a subnetwork with neuronal identity. We reveal an unexpected potential role of Trem2 in the homeostasis of endothelial cells that goes beyond its known functions as a microglial receptor and signaling hub, suggesting an underlying link between immune response and vascular disease in dementia.

Identifying gene network rewiring by combining gene expression and gene mutation data.

Understanding how gene dependency networks rewire between different disease states is an important task in genomic research. Although many computational methods have been proposed to undertake this task via differential network analysis, most of them are designed for a predefined data type. With the development of the high throughput technologies, gene activity measurements can be collected from different aspects (e.g., mRNA expression and DNA mutation). Different data types might share some common characteristics and include certain unique properties. New methods are needed to explore the similarity and difference between differential networks estimated from different data types. In this study, we develop a new differential network inference model which identifies gene network rewiring by combining gene expression and gene mutation data. Similarity and difference between different data types are learned via a group bridge penalty function. Simulation studies have demonstrated that our method consistently outperforms the competing methods. We also apply our method to identify gene network rewiring associated with ovarian cancer platinum resistance. There are certain differential edges common to both data types and some differential edges unique to individual data types. Hub genes in the differential networks inferred by our method play important roles in ovarian cancer drug resistance.

Stromal TRIM28-associated signaling pathway modulation within the colorectal cancer microenvironment

BackgroundStromal gene expression patterns predict patient outcomes in colorectal cancer. TRIM28 is a transcriptional co-repressor that regulates an abundance of genes through the KRAB domain family of transcription factors. We have previously shown that stromal expression of TRIM28 is a marker of disease relapse and poor survival in colorectal cancer. Here, we perform differential epithelium-stroma proteomic network analyses to characterize signaling pathways associated with TRIM28 within the tumor microenvironment.MethodsReverse phase protein arrays were generated from laser capture micro-dissected carcinoma and stromal cells from fresh frozen colorectal cancer tissues. Phosphorylation and total protein levels were measured for 30 cancer-related signaling pathway endpoints. Strength and direction of associations between signaling endpoints were identified using Spearman’s rank-order correlation analysis and compared to TRIM28 levels. Expression status of TRIM28 in tumor epithelium and stromal fibroblasts was assessed using IHC in formalin fixed tissue and the epithelium to stroma protein expression ratio method.ResultsWe found distinct proteomic networks in the epithelial and stromal compartments which were linked to expression levels of TRIM28. Low levels of TRIM28 in tumor stroma (high epithelium: stroma ratio) were found in 10 out of 19 cases. Upon proteomic network analyses, these stromal high ratio cases revealed moderate signaling pathway similarity exemplified by 76 significant Spearman correlations (ρ ≥ 0.75, p ≤ 0.01). Furthermore, low levels of stromal TRIM28 correlated with elevated MDM2 levels in tumor epithelium (p = 0.01) and COX-2 levels in tumor stroma (p = 0.002). Low TRIM28 epithelium to stroma ratios were associated with elevated levels of caspases 3 and 7 in stroma (p = 0.041 and p = 0.036) and an increased signaling pathway similarity in stromal cells with 81 significant Spearman correlations (ρ ≥ 0.75, p ≤ 0.01).ConclusionsBy dissecting TRIM28-associated pathways in stromal fibroblasts and epithelial tumor cells, we performed comprehensive proteomic analyses of molecular networks within the tumor microenvironment. We found modulation of several signaling pathways associated with TRIM28, which may be attributed to the pleiotropic properties of TRIM28 through its translational suppression of the family of KRAB domain transcription factors in tumor stromal compartments.

Niacin action in the atherogenic mixed dyslipidemia of metabolic syndrome: Insights from metabolic biomarker profiling and network analysis

Niacin as an adjunct to statin treatment to reduce cardiovascular risk is questioned. To evaluate interrelationships between the effects of niacin on mixed dyslipidemia and a spectrum of metabolic and inflammatory biomarkers. Obese, nondiabetic, hypertriglyceridemic males (n=19) with low high-density lipoprotein-cholesterol levels received extended-release nicotinic acid for 8weeks. Multiple biomarkers were measured using enzyme-linked immunosorbent assay, enzymatic/absorptiometric, or multiplex biochip assays. Treatment effects were determined for each variable and a differential correlation network created on the basis of univariate correlations between baseline and response to niacin treatment for all pairs of variables. Extended-release niacin treatment favoured normalization of plasma lipid and apolipoprotein profile. Plasma markers of inflammation, hepatic function, cellular adhesion and proliferation, and macrophage phenotype were attenuated; however, insulin resistance increased. Differential network analysis revealed that changes in triglycerides and high-density lipoprotein-cholesterol were closely linked; equally, niacin mediated reductions in total cholesterol, apolipoprotein B, low-density lipoprotein-cholesterol and lipoprotein(a) clustered together, as did homeostatic model assessment of insulin resistance, insulin, and interleukin-6 levels. Two clusters of inflammatory markers were identified, involving (1) intercellular adhesion molecule 1 and high-sensitive C-reactive protein and (2) soluble tumor necrosis factor receptors; and novel clusters involving matrix metallopeptidase 9 and apolipoprotein E, and adiponectin and cystatin C, respectively, were equally revealed. At lower stringency, lipid and insulin resistance clusters were linked; a C-reactive protein-centered cluster linked reduction in apolipoprotein CIII to intercellular adhesion molecule 1, gamma-glutamyltransferase, soluble tumor necrosis factor receptors, and E-selectin. A niacin-mediated trend to normalize atherogenic mixed dyslipidemia was intimately linked to attenuation of biomarkers of inflammation, cell adhesion, hepatic dysfunction and cell proliferation, but to enhanced insulin resistance and plasma homocysteine elevation.

The expression profiling and ontology analysis of non-coding RNAs in dexamethasone induced steatosis in hepatoma cell

Increasing amounts of evidence have indicated that non-coding RNAs (ncRNAs) have important regulatory potential in various biological processes. However, the contribution of ncRNAs, especially long non-coding RNAs (lncRNAs) to drug induced steatosis remain largely unknown. The aim of this study is to investigate miRNA, lncRNA and mRNA expression profiles and their potential roles in the process of drug induced steatosis. Microarray expression profiles of miRNAs, lncRNAs and mRNAs were determined in dexamethasone treated HepG2 cell as well as control cell. Differential expression, pathway and gene network analyses were developed to identify possible functional RNA molecules in dexamethasone induced steatosis. Compared with control HepG2 cell, 652 lncRNAs (528 up-regulated and 124 down-regulated), 655 mRNAs (527 upregulated and 128 down-regulated) and 114 miRNAs (55 miRNAs up-regulated and 59 down-regulated) were differentially expressed in dexamethasone treated HepG2 cell. Pathway analysis showed that the fatty acid biosynthesis, insulin resistance, PPAR signaling pathway, regulation of lipolysis in adipocytes, carbohydrate digestion and absorption, steroid hormone biosynthesis signaling pathways had a close relationship with dexamethasone induced steatosis. 10 highly dysregulated mRNAs and 20 miRNAs, which are closely related to lipid metabolism, were identified and validated by PCR, which followed by ceRNA analysis. CeRNA network analysis identified 5 lipid metabolism related genes, including CYP7A1, CYP11A1, PDK4, ABHD5, ACSL1. It also identified 12 miRNAs (miR-23a-3p, miR-519d-3p, miR-4328, miR-15b-5p etc.) and 177 lncRNAs (ENST00000508884, ENST00000608794, ENST00000568457 etc.). Our results provide a foundation and an expansive view of the roles and mechanisms of ncRNAs in dexamethasone induced steatosis.

DiffGRN: differential gene regulatory network analysis.

Identification of differential gene regulators with significant changes under disparate conditions is essential to understand complex biological mechanism in a disease. Differential Network Analysis (DiNA) examines different biological processes based on gene regulatory networks that represent regulatory interactions between genes with a graph model. While most studies in DiNA have considered correlation-based inference to construct gene regulatory networks from gene expression data due to its intuitive representation and simple implementation, the approach lacks in the representation of causal effects and multivariate effects between genes. In this paper, we propose an approach named Differential Gene Regulatory Network (DiffGRN) that infers differential gene regulation between two groups. We infer gene regulatory networks of two groups using Random LASSO, and then we identify differential gene regulations by the proposed significance test. The advantages of DiffGRN are to capture multivariate effects of genes that regulate a gene simultaneously, to identify causality of gene regulations, and to discover differential gene regulators between regression-based gene regulatory networks. We assessed DiffGRN by simulation experiments and showed its outstanding performance than the current state-of-the-art correlation-based method, DINGO. DiffGRN is applied to gene expression data in asthma. The DiNA with asthma data showed a number of gene regulations, such as ADAM12 and RELB, reported in biological literature.

Potential Breast Anticancer Drug Targets Revealed by Differential Gene Regulatory Network Analysis and Molecular Docking: Neoadjuvant Docetaxel Drug as a Case Study.

Understanding gene-gene interaction and its causal relationship to protein-protein interaction is a viable route for understanding drug action at the genetic level, which is largely hindered by inability to robustly map gene regulatory networks. Here, we use biological prior knowledge of family-to-family gene interactions available in the KEGG database to reveal individual gene-to-gene interaction networks that underlie the gene expression profiles of 2 cell line data sets, sensitive and resistive to neoadjuvant docetaxel breast anticancer drug. Comparison of the topology of the 2 networks revealed that the resistant network is highly connected with 2 large domains of connectivity: one in which the RAF1 and MAP2K2 genes form hubs of connectivity and another in which the RAS gene is highly connected. On the contrary, the sensitive network is highly disrupted with a lower degree of connectivity. We investigated the interactions of the neoadjuvant docetaxel drug with the protein chains encoded by gene-gene interactions that underlie the disruption of the sensitive network topology using protein-protein and drug-protein docking techniques. We found that the sensitive network is likely to be disrupted by interaction of the neoadjuvant docetaxel drug with the DAXX and FGR1 proteins, which is consistent with the observed accumulation of cytoplasmic DAXX and overexpression of FGR1 precursors in cancer cell lines. This indicates that the DAXX and FGR1 proteins could be potential targets for the neoadjuvant docetaxel drug. The work, therefore, provides a new route for understanding the effect of the drug mode of action from the viewpoint of the change in the topology of gene-gene regulatory networks and provides a new avenue for bridging the gap between gene-gene interactions and protein-protein interactions which could have deep implications on mainstream drug development protocols.

IDINGO-integrative differential network analysis in genomics with Shiny application.

MotivationDifferential network analysis is an important way to understand network rewiring involved in disease progression and development. Building differential networks from multiple ‘omics data provides insight into the holistic differences of the interactive system under different patient-specific groups. DINGO was developed to infer group-specific dependencies and build differential networks. However, DINGO and other existing tools are limited to analyze data arising from a single platform, and modeling each of the multiple ‘omics data independently does not account for the hierarchical structure of the data.ResultsWe developed the iDINGO R package to estimate group-specific dependencies and make inferences on the integrative differential networks, considering the biological hierarchy among the platforms. A Shiny application has also been developed to facilitate easier analysis and visualization of results, including integrative differential networks and hub gene identification across platforms.Availability and implementationR package is available on CRAN (https://cran.r-project.org/web/packages/iDINGO) and Shiny application at https://github.com/MinJinHa/iDINGO.Supplementary information Supplementary data are available at Bioinformatics online.

Age and Sex Effects on Plasma Metabolite Association Networks in Healthy Subjects.

In the era of precision medicine, the analysis of simple information like sex and age can increase the potential to better diagnose and treat conditions that occur more frequently in one of the two sexes, present sex-specific symptoms and outcomes, or are characteristic of a specific age group. We present here a study of the association networks constructed from an array of 22 plasma metabolites measured on a cohort of 844 healthy blood donors. Through differential network analysis we show that specific association networks can be associated with sex and age: Different connectivity patterns were observed, suggesting sex-related variability in several metabolic pathways (branched-chain amino acids, ketone bodies, and propanoate metabolism). Reduction in metabolite hub connectivity was also found to be associated with age in both sex groups. Network analysis was complemented with standard univariate and multivariate statistical analysis that revealed age- and sex-specific metabolic signatures. Our results demonstrate that the characterization of metabolite-metabolite association networks is a promising and powerful tool to investigate the human phenotype at a molecular level.

Unique signalling connectivity of FGFR3-TACC3 oncoprotein revealed by quantitative phosphoproteomics and differential network analysis.

The FGFR3-TACC3 fusion is an oncogenic driver in diverse malignancies, including bladder cancer, characterized by upregulated tyrosine kinase activity. To gain insights into distinct properties of FGFR3-TACC3 down-stream signalling, we utilised telomerase-immortalised normal human urothelial cell lines expressing either the fusion or wild-type FGFR3 (isoform IIIb) for subsequent quantitative proteomics and network analysis. Cellular lysates were chemically labelled with isobaric tandem mass tag reagents and, after phosphopeptide enrichment, liquid chromatography-high mass accuracy tandem mass spectrometry (LC-MS/MS) was used for peptide identification and quantification. Comparison of data from the two cell lines under non-stimulated and FGF1 stimulated conditions and of data representing physiological stimulation of FGFR3 identified about 200 regulated phosphosites. The identified phosphoproteins and quantified phosphosites were further analysed in the context of functional biological networks by inferring kinase-substrate interactions, mapping these to a comprehensive human signalling interaction network, filtering based on tissue-expression profiles and applying disease module detection and pathway enrichment methods. Analysis of our phosphoproteomics data using these bioinformatics methods combined into a new protocol—Disease Relevant Analysis of Genes On Networks (DRAGON)—allowed us to tease apart pathways differentially involved in FGFR3-TACC3 signalling in comparison to wild-type FGFR3 and to investigate their local phospho-signalling context. We highlight 9 pathways significantly regulated only in the cell line expressing FGFR3-TACC3 fusion and 5 pathways regulated only by stimulation of the wild-type FGFR3. Pathways differentially linked to FGFR3-TACC3 fusion include those related to chaperone activation and stress response and to regulation of TP53 expression and degradation that could contribute to development and maintenance of the cancer phenotype.

XMWAS: a data-driven integration and differential network analysis tool.

Integrative omics is a central component of most systems biology studies. Computational methods are required for extracting meaningful relationships across different omics layers. Various tools have been developed to facilitate integration of paired heterogenous omics data; however most existing tools allow integration of only two omics datasets. Furthermore, existing data integration tools do not incorporate additional steps of identifying sub-networks or communities of highly connected entities and evaluating the topology of the integrative network under different conditions. Here we present xMWAS, a software for data integration, network visualization, clustering, and differential network analysis of data from biochemical and phenotypic assays, and two or more omics platforms. https://kuppal.shinyapps.io/xmwas (Online) and https://github.com/kuppal2/xMWAS/ (R). kuppal2@emory.edu. Supplementary data are available at Bioinformatics online.

Differential network analysis via lasso penalized D-trace loss

SummaryBiological networks often change under different environmental and genetic conditions. In this paper, we model network change as the difference of two precision matrices and propose a novel loss function called the D-trace loss, which allows us to directly estimate the precision matrix difference without attempting to estimate the precision matrices themselves. Under a new irrepresentability condition, we show that the D-trace loss function with the lasso penalty can yield consistent estimators in high-dimensional settings if the difference network is sparse. A very efficient algorithm is developed based on the alternating direction method of multipliers to minimize the penalized loss function. Simulation studies and a real-data analysis show that the proposed method outperforms other methods.

Differential gene regulatory networks in development and disease.

Gene regulatory networks, in which differential expression of regulator genes induce differential expression of their target genes, underlie diverse biological processes such as embryonic development, organ formation and disease pathogenesis. An archetypical systems biology approach to mapping these networks involves the combined application of (1) high-throughput sequencing-based transcriptome profiling (RNA-seq) of biopsies under diverse network perturbations and (2) network inference based on gene-gene expression correlation analysis. The comparative analysis of such correlation networks across cell types or states, differential correlation network analysis, can identify specific molecular signatures and functional modules that underlie the state transition or have context-specific function. Here, we review the basic concepts of network biology and correlation network inference, and the prevailing methods for differential analysis of correlation networks. We discuss applications of gene expression network analysis in the context of embryonic development, cancer, and congenital diseases.

Gene Network Construction from Microarray Data Identifies a Key Network Module and Several Candidate Hub Genes in Age-Associated Spatial Learning Impairment.

As humans age many suffer from a decrease in normal brain functions including spatial learning impairments. This study aimed to better understand the molecular mechanisms in age-associated spatial learning impairment (ASLI). We used a mathematical modeling approach implemented in Weighted Gene Co-expression Network Analysis (WGCNA) to create and compare gene network models of young (learning unimpaired) and aged (predominantly learning impaired) brains from a set of exploratory datasets in rats in the context of ASLI. The major goal was to overcome some of the limitations previously observed in the traditional meta- and pathway analysis using these data, and identify novel ASLI related genes and their networks based on co-expression relationship of genes. This analysis identified a set of network modules in the young, each of which is highly enriched with genes functioning in broad but distinct GO functional categories or biological pathways. Interestingly, the analysis pointed to a single module that was highly enriched with genes functioning in “learning and memory” related functions and pathways. Subsequent differential network analysis of this “learning and memory” module in the aged (predominantly learning impaired) rats compared to the young learning unimpaired rats allowed us to identify a set of novel ASLI candidate hub genes. Some of these genes show significant repeatability in networks generated from independent young and aged validation datasets. These hub genes are highly co-expressed with other genes in the network, which not only show differential expression but also differential co-expression and differential connectivity across age and learning impairment. The known function of these hub genes indicate that they play key roles in critical pathways, including kinase and phosphatase signaling, in functions related to various ion channels, and in maintaining neuronal integrity relating to synaptic plasticity and memory formation. Taken together, they provide a new insight and generate new hypotheses into the molecular mechanisms responsible for age associated learning impairment, including spatial learning.

Network analysis of patient flow in two UK acute care hospitals identifies key sub-networks for A&E performance.

The topology of the patient flow network in a hospital is complex, comprising hundreds of overlapping patient journeys, and is a determinant of operational efficiency. To understand the network architecture of patient flow, we performed a data-driven network analysis of patient flow through two acute hospital sites of King’s College Hospital NHS Foundation Trust. Administration databases were queried for all intra-hospital patient transfers in an 18-month period and modelled as a dynamic weighted directed graph. A ‘core’ subnetwork containing only 13–17% of all edges channelled 83–90% of the patient flow, while an ‘ephemeral’ network constituted the remainder. Unsupervised cluster analysis and differential network analysis identified sub-networks where traffic is most associated with A&E performance. Increased flow to clinical decision units was associated with the best A&E performance in both sites. The component analysis also detected a weekend effect on patient transfers which was not associated with performance. We have performed the first data-driven hypothesis-free analysis of patient flow which can enhance understanding of whole healthcare systems. Such analysis can drive transformation in healthcare as it has in industries such as manufacturing.