Pillars of biology: Boolean modeling of gene-regulatory networks.

Boolean modeling is a mathematical modeling framework used for defining and studying gene-regulatory networks (GRNs). It serves as a means to develop mechanistic models, offering insights into the trajectories and dynamic properties of GRNs. In this review, I delve into seminal papers published in the Journal of Theoretical Biology that have spearheaded this field. Additionally, I explore the application of these modeling methods in the current era of data-intensive science.

Discrete-state models identify pathway specific B cell states across diseases and infections at single-cell resolution.

Oxygen (O2) regulated pathways modulate B cell activation, migration and proliferation during infection, vaccination, and other diseases. Modeling these pathways in health and disease is critical to understand B cell states and ways to mediate them. To characterize B cells by their activation of O2 regulated pathways we develop pathway specific discrete state models using previously published single-cell RNA-sequencing (scRNA-seq) datasets from isolated B cells. Specifically, Single Cell Boolean Omics Network Invariant-Time Analysis (scBONITA) was used to infer logic gates for known pathway topologies. The simplest inferred set of logic gates that maximized the number of "OR" interactions between genes was used to simulate B cell networks involved in oxygen sensing until they reached steady network states (attractors). By focusing on the attractors that best represented sequenced cells, we identified genes critical in determining pathway specific cellular states that corresponded to diseased and healthy B cell phenotypes. Specifically, we investigate the transendothelial migration, regulation of actin cytoskeleton, HIF1A, and Citrate Cycle pathways. Our analysis revealed attractors that resembled the state of B cell exhaustion in HIV+ patients as well as attractors that promoted anerobic metabolism, angiogenesis, and tumorigenesis in breast cancer patients, which were eliminated after neoadjuvant chemotherapy (NACT). Finally, we investigated the attractors to which the Azimuth-annotated B cells mapped and found that attractors resembling B cells from HIV+ patients encompassed a significantly larger number of atypical memory B cells than HIV- attractors. Meanwhile, attractors resembling B cells from breast cancer patients post NACT encompassed a reduced number of atypical memory B cells compared to pre-NACT attractors.

Executable Network Models of Integrated Multiomics Data.

Multiomics profiling provides a holistic picture of a condition being examined and captures the complexity of signaling events, beginning from the original cause (environmental or genetic), to downstream functional changes at multiple molecular layers. Pathway enrichment analysis has been used with multiomics data sets to characterize signaling mechanisms. However, technical and biological variability between these layered data limit an integrative computational analyses. We present a Boolean network-based method, multiomics Boolean Omics Network Invariant-Time Analysis (mBONITA), to integrate omics data sets that quantify multiple molecular layers. mBONITA utilizes prior knowledge networks to perform topology-based pathway analysis. In addition, mBONITA identifies genes that are consistently modulated across molecular measurements by combining observed fold-changes and variance, with a measure of node (i.e., gene or protein) influence over signaling, and a measure of the strength of evidence for that gene across data sets. We used mBONITA to integrate multiomics data sets from RAMOS B cells treated with the immunosuppressant drug cyclosporine A under varying O2 tensions to identify pathways involved in hypoxia-mediated chemotaxis. We compare mBONITA's performance with 6 other pathway analysis methods designed for multiomics data and show that mBONITA identifies a set of pathways with evidence of modulation across all omics layers. mBONITA is freely available at https://github.com/Thakar-Lab/mBONITA.

WikiNetworks: translating manually created biological pathways for topological analysis.

SUMMARY: WikiPathways is a database of 2979 biological pathways across 31 species created using the drawing software PathVisio. Many of these pathways are not directly usable for network-based topological analyses due to differences in curation styles and drawings. We developed the WikiNetworks package to standardize and construct directed networks by combining geometric information and manual annotations from WikiPathways. WikiNetworks performs significantly better than existing tools. This enables the use of high-quality WikiPathways resource for network-based topological analysis of high-throughput data. AVAILABILITY AND IMPLEMENTATION: WikiNetworks is written in Python3 and is available on github.com/Thakar-Lab/wikinetworks and on PyPI. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

SysMod: the ISCB community for data-driven computational modelling and multi-scale analysis of biological systems.

Computational models of biological systems can exploit a broad range of rapidly developing approaches, including novel experimental approaches, bioinformatics data analysis, emerging modelling paradigms, data standards and algorithms. A discussion about the most recent advances among experts from various domains is crucial to foster data-driven computational modelling and its growing use in assessing and predicting the behaviour of biological systems. Intending to encourage the development of tools, approaches and predictive models, and to deepen our understanding of biological systems, the Community of Special Interest (COSI) was launched in Computational Modelling of Biological Systems (SysMod) in 2016. SysMod's main activity is an annual meeting at the Intelligent Systems for Molecular Biology (ISMB) conference, which brings together computer scientists, biologists, mathematicians, engineers, computational and systems biologists. In the five years since its inception, SysMod has evolved into a dynamic and expanding community, as the increasing number of contributions and participants illustrate. SysMod maintains several online resources to facilitate interaction among the community members, including an online forum, a calendar of relevant meetings and a YouTube channel with talks and lectures of interest for the modelling community. For more than half a decade, the growing interest in computational systems modelling and multi-scale data integration has inspired and supported the SysMod community. Its members get progressively more involved and actively contribute to the annual COSI meeting and several related community workshops and meetings, focusing on specific topics, including particular techniques for computational modelling or standardisation efforts.

A systems genomics approach uncovers molecular associates of RSV severity.

Respiratory syncytial virus (RSV) infection results in millions of hospitalizations and thousands of deaths each year. Variations in the adaptive and innate immune response appear to be associated with RSV severity. To investigate the host response to RSV infection in infants, we performed a systems-level study of RSV pathophysiology, incorporating high-throughput measurements of the peripheral innate and adaptive immune systems and the airway epithelium and microbiota. We implemented a novel multi-omic data integration method based on multilayered principal component analysis, penalized regression, and feature weight back-propagation, which enabled us to identify cellular pathways associated with RSV severity. In both airway and immune cells, we found an association between RSV severity and activation of pathways controlling Th17 and acute phase response signaling, as well as inhibition of B cell receptor signaling. Dysregulation of both the humoral and mucosal response to RSV may play a critical role in determining illness severity.

Infant gut microbiome is enriched with Bifidobacterium longum ssp. infantis in Old Order Mennonites with traditional farming lifestyle.

BACKGROUND: Growing up on traditional, single-family farms is associated with protection against asthma in school age, but the mechanisms against early manifestations of atopic disease are largely unknown. We sought determine the gut microbiome and metabolome composition in rural Old Order Mennonite (OOM) infants at low risk and Rochester, NY urban/suburban infants at high risk for atopic diseases. METHODS: In a cohort of 65 OOM and 39 Rochester mother-infant pairs, 101 infant stool and 61 human milk samples were assessed by 16S rRNA gene sequencing for microbiome composition and qPCR to quantify Bifidobacterium spp. and B. longum ssp. infantis (B. infantis), a consumer of human milk oligosaccharides (HMOs). Fatty acids (FAs) were analyzed in 34 stool and human 24 milk samples. Diagnoses and symptoms of atopic diseases by 3 years of age were assessed by telephone. RESULTS: At a median age of 2 months, stool was enriched with Bifidobacteriaceae, Clostridiaceae, and Aerococcaceae in the OOM compared with Rochester infants. B. infantis was more abundant (p < .001) and prevalent, detected in 70% of OOM compared with 21% of Rochester infants (p < .001). Stool colonized with B. infantis had higher levels of lactate and several medium- to long/odd-chain FAs. In contrast, paired human milk was enriched with a distinct set of FAs including butyrate. Atopic diseases were reported in 6.5% of OOM and 35% of Rochester children (p < .001). CONCLUSION: A high rate of B. infantis colonization, similar to that seen in developing countries, is found in the OOM at low risk for atopic diseases.

Cyclosporine a directly affects human and mouse b cell migration in vitro by disrupting a hIF-1 αdependent, o2 sensing, molecular switch.

BACKGROUND: Hypoxia is a potent molecular signal for cellular metabolism, mitochondrial function, and migration. Conditions of low oxygen tension trigger regulatory cascades mediated via the highly conserved HIF-1 α post-translational modification system. In the adaptive immune response, B cells (Bc) are activated and differentiate under hypoxic conditions within lymph node germinal centers, and subsequently migrate to other compartments. During migration, they traverse through changing oxygen levels, ranging from 1-5% in the lymph node to 5-13% in the peripheral blood. Interestingly, the calcineurin inhibitor cyclosporine A is known to stimulate prolyl hydroxylase activity, resulting in HIF-1 α destabilization and may alter Bc responses directly. Over 60% of patients taking calcineurin immunosuppressant medications have hypo-gammaglobulinemia and poor vaccine responses, putting them at high risk of infection with significantly increased morbidity and mortality. RESULTS: We demonstrate that O 2 tension is a previously unrecognized Bc regulatory switch, altering CXCR4 and CXCR5 chemokine receptor signaling in activated Bc through HIF-1 α expression, and controlling critical aspects of Bc migration. Our data demonstrate that calcineurin inhibition hinders this O 2 regulatory switch in primary human Bc. CONCLUSION: This previously unrecognized effect of calcineurin inhibition directly on human Bc has significant and direct clinical implications.

Executable pathway analysis using ensemble discrete-state modeling for large-scale data.

Pathway analysis is widely used to gain mechanistic insights from high-throughput omics data. However, most existing methods do not consider signal integration represented by pathway topology, resulting in enrichment of convergent pathways when downstream genes are modulated. Incorporation of signal flow and integration in pathway analysis could rank the pathways based on modulation in key regulatory genes. This implementation can be facilitated for large-scale data by discrete state network modeling due to simplicity in parameterization. Here, we model cellular heterogeneity using discrete state dynamics and measure pathway activities in cross-sectional data. We introduce a new algorithm, Boolean Omics Network Invariant-Time Analysis (BONITA), for signal propagation, signal integration, and pathway analysis. Our signal propagation approach models heterogeneity in transcriptomic data as arising from intercellular heterogeneity rather than intracellular stochasticity, and propagates binary signals repeatedly across networks. Logic rules defining signal integration are inferred by genetic algorithm and are refined by local search. The rules determine the impact of each node in a pathway, which is used to score the probability of the pathway's modulation by chance. We have comprehensively tested BONITA for application to transcriptomics data from translational studies. Comparison with state-of-the-art pathway analysis methods shows that BONITA has higher sensitivity at lower levels of source node modulation and similar sensitivity at higher levels of source node modulation. Application of BONITA pathway analysis to previously validated RNA-sequencing studies identifies additional relevant pathways in in-vitro human cell line experiments and in-vivo infant studies. Additionally, BONITA successfully detected modulation of disease specific pathways when comparing relevant RNA-sequencing data with healthy controls. Most interestingly, the two highest impact score nodes identified by BONITA included known drug targets. Thus, BONITA is a powerful approach to prioritize not only pathways but also specific mechanistic role of genes compared to existing methods. BONITA is available at: https://github.com/thakar-lab/BONITA.

IL-1ß-driven amyloid plaque clearance is associated with an expansion of transcriptionally reprogrammed microglia.

BACKGROUND: Neuroinflammation is thought to contribute to the pathogenesis of Alzheimer's disease (AD), yet numerous studies have demonstrated a beneficial role for neuroinflammation in amyloid plaque clearance. We have previously shown that sustained expression of IL-1ß in the hippocampus of APP/PS1 mice decreases amyloid plaque burden independent of recruited CCR2+ myeloid cells, suggesting resident microglia as the main phagocytic effectors of IL-1ß-induced plaque clearance. To date, however, the mechanisms of IL-1ß-induced plaque clearance remain poorly understood. METHODS: To determine whether microglia are involved in IL-1ß-induced plaque clearance, APP/PS1 mice induced to express mature human IL-1ß in the hippocampus via adenoviral transduction were treated with the Aß fluorescent probe methoxy-X04 (MX04) and microglial internalization of fibrillar Aß (fAß) was analyzed by flow cytometry and immunohistochemistry. To assess microglial proliferation, APP/PS1 mice transduced with IL-1ß or control were injected intraperitoneally with BrdU and hippocampal tissue was analyzed by flow cytometry. RNAseq analysis was conducted on microglia FACS sorted from the hippocampus of control or IL-1ß-treated APP/PS1 mice. These microglia were also sorted based on MX04 labeling (MX04+ and MX04- microglia). RESULTS: Resident microglia (CD45loCD11b+) constituted > 70% of the MX04+ cells in both Phe- and IL-1ß-treated conditions, and < 15% of MX04+ cells were recruited myeloid cells (CD45hiCD11b+). However, IL-1ß treatment did not augment the percentage of MX04+ microglia nor the quantity of fAß internalized by individual microglia. Instead, IL-1ß increased the total number of MX04+ microglia in the hippocampus due to IL-1ß-induced proliferation. In addition, transcriptomic analyses revealed that IL-1ß treatment was associated with large-scale changes in the expression of genes related to immune responses, proliferation, and cytokine signaling. CONCLUSIONS: These studies show that IL-1ß overexpression early in amyloid pathogenesis induces a change in the microglial gene expression profile and an expansion of microglial cells that facilitates Aß plaque clearance.

FUNNEL-GSEA: FUNctioNal ELastic-net regression in time-course gene set enrichment analysis.

MOTIVATION: Gene set enrichment analyses (GSEAs) are widely used in genomic research to identify underlying biological mechanisms (defined by the gene sets), such as Gene Ontology terms and molecular pathways. There are two caveats in the currently available methods: (i) they are typically designed for group comparisons or regression analyses, which do not utilize temporal information efficiently in time-series of transcriptomics measurements; and (ii) genes overlapping in multiple molecular pathways are considered multiple times in hypothesis testing. RESULTS: We propose an inferential framework for GSEA based on functional data analysis, which utilizes the temporal information based on functional principal component analysis, and disentangles the effects of overlapping genes by a functional extension of the elastic-net regression. Furthermore, the hypothesis testing for the gene sets is performed by an extension of Mann-Whitney U test which is based on weighted rank sums computed from correlated observations. By using both simulated datasets and a large-scale time-course gene expression data on human influenza infection, we demonstrate that our method has uniformly better receiver operating characteristic curves, and identifies more pathways relevant to immune-response to human influenza infection than the competing approaches. AVAILABILITY AND IMPLEMENTATION: The methods are implemented in R package FUNNEL, freely and publicly available at: https://github.com/yunzhang813/FUNNEL-GSEA-R-Package . CONTACT: xing_qiu@urmc.rochester.edu or juilee_thakar@urmc.rochester.edu. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Association of Dynamic Changes in the CD4 T-Cell Transcriptome With Disease Severity During Primary Respiratory Syncytial Virus Infection in Young Infants.

Background: Nearly all children are infected with respiratory syncytial virus (RSV) within the first 2 years of life, with a minority developing severe disease (1%-3% hospitalized). We hypothesized that an assessment of the adaptive immune system, using CD4+ T-lymphocyte transcriptomics, would identify gene expression correlates of disease severity. Methods: Infants infected with RSV representing extremes of clinical severity were studied. Mild illness (n = 23) was defined as a respiratory rate (RR) < 55 and room air oxygen saturation (SaO2) ≥ 97%, and severe illness (n = 23) was defined as RR ≥ 65 and SaO2 ≤ 92%. RNA from fresh, sort-purified CD4+ T cells was assessed by RNA sequencing. Results: Gestational age, age at illness onset, exposure to environmental tobacco smoke, bacterial colonization, and breastfeeding were associated (adjusted P < .05) with disease severity. RNA sequencing analysis reliably measured approximately 60% of the genome. Severity of RSV illness had the greatest effect size upon CD4 T-cell gene expression. Pathway analysis identified correlates of severity, including JAK/STAT, prolactin, and interleukin 9 signaling. We also identified genes and pathways associated with timing of symptoms and RSV group (A/B). Conclusions: These data suggest fundamental changes in adaptive immune cell phenotypes may be associated with RSV clinical severity.

Meta-analysis of cell- specific transcriptomic data using fuzzy c-means clustering discovers versatile viral responsive genes.

BACKGROUND: Despite advances in the gene-set enrichment analysis methods; inadequate definitions of gene-sets cause a major limitation in the discovery of novel biological processes from the transcriptomic datasets. Typically, gene-sets are obtained from publicly available pathway databases, which contain generalized definitions frequently derived by manual curation. Recently unsupervised clustering algorithms have been proposed to identify gene-sets from transcriptomics datasets deposited in public domain. These data-driven definitions of the gene-sets can be context-specific revealing novel biological mechanisms. However, the previously proposed algorithms for identification of data-driven gene-sets are based on hard clustering which do not allow overlap across clusters, a characteristic that is predominantly observed across biological pathways. RESULTS: We developed a pipeline using fuzzy-C-means (FCM) soft clustering approach to identify gene-sets which recapitulates topological characteristics of biological pathways. Specifically, we apply our pipeline to derive gene-sets from transcriptomic data measuring response of monocyte derived dendritic cells and A549 epithelial cells to influenza infections. Our approach apply Ward's method for the selection of initial conditions, optimize parameters of FCM algorithm for human cell-specific transcriptomic data and identify robust gene-sets along with versatile viral responsive genes. CONCLUSION: We validate our gene-sets and demonstrate that by identifying genes associated with multiple gene-sets, FCM clustering algorithm significantly improves interpretation of transcriptomic data facilitating investigation of novel biological processes by leveraging on transcriptomic data available in the public domain. We develop an interactive 'Fuzzy Inference of Gene-sets (FIGS)' package (GitHub: https://github.com/Thakar-Lab/FIGS ) to facilitate use of of pipeline. Future extension of FIGS across different immune cell-types will improve mechanistic investigation followed by high-throughput omics studies.

Optimized logic rules reveal interferon-γ-induced modes regulated by histone deacetylases and protein tyrosine phosphatases.

The pro-inflammatory cytokine interferon-γ (IFN-γ) is critical for activating innate and adaptive immunity against tumours and intracellular pathogens. Interferon-γ is secreted at the fetal-maternal interface in pregnant women and mice. The outer layer of the placenta in contact with maternal blood is composed of semi-allogeneic trophoblast cells, which constitute the fetal component of the fetal-maternal interface. The simultaneous presence of pro-inflammatory IFN-γ and trophoblast cells at the fetal-maternal interface appears to represent an immunological paradox, for trophoblastic responses to IFN-γ could potentially lead to activation of maternal immunity and subsequent attack of the placenta. However, our previous studies demonstrate that IFN-γ responsive gene (IRG) expression is negatively regulated in human and mouse trophoblast cells. In human cytotrophoblast and trophoblast-derived choriocarcinoma cells, janus kinase signalling is blocked by protein tyrosine phosphatases (PTPs), whereas in mouse trophoblast, histone deacetylases (HDACs) inhibit IRG expression. Here, we used genome-wide transcriptional profiling to investigate the collective roles of PTPs and HDACs on regulation of IRG expression in human choriocarcinoma cells. Logic-rules were optimized to derive regulatory modes governing gene expression patterns observed upon different combinations of treatment with PTP and HDAC inhibitors. The results demonstrate that IRGs can be divided into several categories in human choriocarcinoma cells, each of which is subject to distinct mechanisms of repression. Hence, the regulatory modes identified in this study suggest that human trophoblast and choriocarcinoma cells may evade the potentially deleterious consequences of exposure to IFN-γ by using several overlapping mechanisms to block IRG expression.

Human Dendritic Cell Response Signatures Distinguish 1918, Pandemic, and Seasonal H1N1 Influenza Viruses.

UNLABELLED: Influenza viruses continue to present global threats to human health. Antigenic drift and shift, genetic reassortment, and cross-species transmission generate new strains with differences in epidemiology and clinical severity. We compared the temporal transcriptional responses of human dendritic cells (DC) to infection with two pandemic (A/Brevig Mission/1/1918, A/California/4/2009) and two seasonal (A/New Caledonia/20/1999, A/Texas/36/1991) H1N1 influenza viruses. Strain-specific response differences included stronger activation of NF-κB following infection with A/New Caledonia/20/1999 and a unique cluster of genes expressed following infection with A/Brevig Mission/1/1918. A common antiviral program showing strain-specific timing was identified in the early DC response and found to correspond with reported transcript changes in blood during symptomatic human influenza virus infection. Comparison of the global responses to the seasonal and pandemic strains showed that a dramatic divergence occurred after 4 h, with only the seasonal strains inducing widespread mRNA loss. IMPORTANCE: Continuously evolving influenza viruses present a global threat to human health; however, these host responses display strain-dependent differences that are incompletely understood. Thus, we conducted a detailed comparative study assessing the immune responses of human DC to infection with two pandemic and two seasonal H1N1 influenza strains. We identified in the immune response to viral infection both common and strain-specific features. Among the stain-specific elements were a time shift of the interferon-stimulated gene response, selective induction of NF-κB signaling by one of the seasonal strains, and massive RNA degradation as early as 4 h postinfection by the seasonal, but not the pandemic, viruses. These findings illuminate new aspects of the distinct differences in the immune responses to pandemic and seasonal influenza viruses.

Competitive SWIFT cluster templates enhance detection of aging changes.

Clustering-based algorithms for automated analysis of flow cytometry datasets have achieved more efficient and objective analysis than manual processing. Clustering organizes flow cytometry data into subpopulations with substantially homogenous characteristics but does not directly address the important problem of identifying the salient differences in subpopulations between subjects and groups. Here, we address this problem by augmenting SWIFT--a mixture model based clustering algorithm reported previously. First, we show that SWIFT clustering using a "template" mixture model, in which all subpopulations are represented, identifies small differences in cell numbers per subpopulation between samples. Second, we demonstrate that resolution of inter-sample differences is increased by "competition" wherein a joint model is formed by combining the mixture model templates obtained from different groups. In the joint model, clusters from individual groups compete for the assignment of cells, sharpening differences between samples, particularly differences representing subpopulation shifts that are masked under clustering with a single template model. The benefit of competition was demonstrated first with a semisynthetic dataset obtained by deliberately shifting a known subpopulation within an actual flow cytometry sample. Single templates correctly identified changes in the number of cells in the subpopulation, but only the competition method detected small changes in median fluorescence. In further validation studies, competition identified a larger number of significantly altered subpopulations between young and elderly subjects. This enrichment was specific, because competition between templates from consensus male and female samples did not improve the detection of age-related differences. Several changes between the young and elderly identified by SWIFT template competition were consistent with known alterations in the elderly, and additional altered subpopulations were also identified. Alternative algorithms detected far fewer significantly altered clusters. Thus SWIFT template competition is a powerful approach to sharpen comparisons between selected groups in flow cytometry datasets.

GESPA: classifying nsSNPs to predict disease association.

BACKGROUND: Non-synonymous single nucleotide polymorphisms (nsSNPs) are the most common DNA sequence variation associated with disease in humans. Thus determining the clinical significance of each nsSNP is of great importance. Potential detrimental nsSNPs may be identified by genetic association studies or by functional analysis in the laboratory, both of which are expensive and time consuming. Existing computational methods lack accuracy and features to facilitate nsSNP classification for clinical use. We developed the GESPA (GEnomic Single nucleotide Polymorphism Analyzer) program to predict the pathogenicity and disease phenotype of nsSNPs. RESULTS: GESPA is a user-friendly software package for classifying disease association of nsSNPs. It allows flexibility in acceptable input formats and predicts the pathogenicity of a given nsSNP by assessing the conservation of amino acids in orthologs and paralogs and supplementing this information with data from medical literature. The development and testing of GESPA was performed using the humsavar, ClinVar and humvar datasets. Additionally, GESPA also predicts the disease phenotype associated with a nsSNP with high accuracy, a feature unavailable in existing software. GESPA's overall accuracy exceeds existing computational methods for predicting nsSNP pathogenicity. The usability of GESPA is enhanced by fast SQL-based cloud storage and retrieval of data. CONCLUSIONS: GESPA is a novel bioinformatics tool to determine the pathogenicity and phenotypes of nsSNPs. We anticipate that GESPA will become a useful clinical framework for predicting the disease association of nsSNPs. The program, executable jar file, source code, GPL 3.0 license, user guide, and test data with instructions are available at http://sourceforge.net/projects/gespa.

Comparative analysis of anti-viral transcriptomics reveals novel effects of influenza immune antagonism.

BACKGROUND: Comparative analysis of genome-wide expression profiles are increasingly being used to study virus-specific host interactions. In order to gain mechanistic insights, gene expression profiles can be combined with information on DNA-binding sites of transcription factors to detect transcription factor activity (by analysis of target gene sets) during viral infections. Here, we apply this approach to study mechanisms of immune antagonism elicited by Influenza A virus (New Caledonia/20/1999) by comparing the transcriptional response with the non-pathogenic Newcastle disease virus (NDV), which lacks human immune antagonism. RESULTS: Existing gene set approaches do not quantify activity in a way that can be statistically compared between responses. We thus developed a new method for Bayesian Estimation of Transcription factor Activity (BETA) that allows for such quantification and comparative analysis across multiple responses. BETA predicted decreased ISGF3 activity during influenza A infection of human dendritic cells (reflected in lower expression of Interferon Stimulated Genes, ISGs). This prediction was confirmed through a combination of mathematical modeling and experiments at different multiplicities of infection to show that ISGs were specifically blocked in infected cells. Suppression of the transcription factor SATB1 was also predicted as a novel effect of influenza-mediated immune antagonism, and validated experimentally. CONCLUSIONS: Comparative analysis of genome-wide transcriptional profiles can reveal new effects of viral immune antagonism. We have developed a computational framework (BETA) that enables quantitative comparative analysis of transcription factor activities. This method will aid future studies to identify mechanistic differences in the host-pathogen interactions. Application of BETA to genome-wide transcriptional profiling data from human DCs identified SATB1 as a novel effect of influenza antagonism.

Quantitative set analysis for gene expression: a method to quantify gene set differential expression including gene-gene correlations.

Enrichment analysis of gene sets is a popular approach that provides a functional interpretation of genome-wide expression data. Existing tests are affected by inter-gene correlations, resulting in a high Type I error. The most widely used test, Gene Set Enrichment Analysis, relies on computationally intensive permutations of sample labels to generate a null distribution that preserves gene-gene correlations. A more recent approach, CAMERA, attempts to correct for these correlations by estimating a variance inflation factor directly from the data. Although these methods generate P-values for detecting gene set activity, they are unable to produce confidence intervals or allow for post hoc comparisons. We have developed a new computational framework for Quantitative Set Analysis of Gene Expression (QuSAGE). QuSAGE accounts for inter-gene correlations, improves the estimation of the variance inflation factor and, rather than evaluating the deviation from a null hypothesis with a P-value, it quantifies gene-set activity with a complete probability density function. From this probability density function, P-values and confidence intervals can be extracted and post hoc analysis can be carried out while maintaining statistical traceability. Compared with Gene Set Enrichment Analysis and CAMERA, QuSAGE exhibits better sensitivity and specificity on real data profiling the response to interferon therapy (in chronic Hepatitis C virus patients) and Influenza A virus infection. QuSAGE is available as an R package, which includes the core functions for the method as well as functions to plot and visualize the results.