Improved data quality and statistical power of trial-level event-related potentials with Bayesian random-shift Gaussian processes

Studies of cognitive processes via electroencephalogram (EEG) recordings often analyze group-level event-related potentials (ERPs) averaged over multiple subjects and trials. This averaging procedure can obscure scientifically relevant variability across subjects and trials, but has been necessary due to the difficulties posed by inference of trial-level ERPs. We introduce the Bayesian Random Phase-Amplitude Gaussian Process (RPAGP) model, for inference of trial-level amplitude, latency, and ERP waveforms. We apply RPAGP to data from a study of ERP responses to emotionally arousing images. The model estimates of trial-specific signals are shown to greatly improve statistical power in detecting significant differences in experimental conditions compared to existing methods. Our results suggest that replacing the observed data with the de-noised RPAGP predictions can potentially improve the sensitivity and accuracy of many of the existing ERP analysis pipelines.

Diminished Adaptation, Satisfaction, and Neural Responses to Advantageous Social Signals in Anorexia Nervosa and Bulimia Nervosa

BACKGROUND

Development and recurrence of 2 eating disorders (EDs), anorexia nervosa and bulimia nervosa, are frequently associated with environmental stressors. Neurobehavioral responses to social learning signals were evaluated in both EDs.

METHODS

Women with anorexia nervosa (n = 25), women with bulimia nervosa (n = 30), or healthy comparison women (n = 38) played a neuroeconomic game in which the norm shifted, generating social learning signals (norm prediction errors [NPEs]) during a functional magnetic resonance imaging scan. A Bayesian logistic regression model examined how the probability of offer acceptance depended on cohort, block, and NPEs. Rejection rates, emotion ratings, and neural responses to NPEs were compared across groups.

RESULTS

Relative to the comparison group, both ED cohorts showed less adaptation (p = .028, ηp2 = 0.060), and advantageous signals (positive NPEs) led to higher rejection rates (p = .014, ηp2 = 0.077) and less positive emotion ratings (p = .004, ηp2 = 0.111). Advantageous signals increased neural activations in the orbitofrontal cortex for the comparison group but not for women with anorexia nervosa (p = .018, d = 0.655) or bulimia nervosa (p = .043, d = 0.527). More severe ED symptoms were associated with decreased activation of dorsomedial prefrontal cortex for advantageous signals.

CONCLUSIONS

Diminished neural processing of advantageous social signals and impaired norm adaptation were observed in both anorexia nervosa and bulimia nervosa, while no differences were found for disadvantageous social signals. Development of neurocognitive interventions to increase responsivity to advantageous social signals could augment current treatments, potentially leading to improved clinical outcomes for EDs.

Noradrenaline tracks emotional modulation of attention in human amygdala

The noradrenaline (NA) system is one of the brain's major neuromodulatory systems; it originates in a small midbrain nucleus, the locus coeruleus (LC), and projects widely throughout the brain.1,2 The LC-NA system is believed to regulate arousal and attention3,4 and is a pharmacological target in multiple clinical conditions.5,6,7 Yet our understanding of its role in health and disease has been impeded by a lack of direct recordings in humans. Here, we address this problem by showing that electrochemical estimates of sub-second NA dynamics can be obtained using clinical depth electrodes implanted for epilepsy monitoring. We made these recordings in the amygdala, an evolutionarily ancient structure that supports emotional processing8,9 and receives dense LC-NA projections,10 while patients (n = 3) performed a visual affective oddball task. The task was designed to induce different cognitive states, with the oddball stimuli involving emotionally evocative images,11 which varied in terms of arousal (low versus high) and valence (negative versus positive). Consistent with theory, the NA estimates tracked the emotional modulation of attention, with a stronger oddball response in a high-arousal state. Parallel estimates of pupil dilation, a common behavioral proxy for LC-NA activity,12 supported a hypothesis that pupil-NA coupling changes with cognitive state,13,14 with the pupil and NA estimates being positively correlated for oddball stimuli in a high-arousal but not a low-arousal state. Our study provides proof of concept that neuromodulator monitoring is now possible using depth electrodes in standard clinical use.

Bayesian continuous-time hidden Markov models with covariate selection for intensive longitudinal data with measurement error

Intensive longitudinal data collected with ecological momentary assessment methods capture information on participants' behaviors, feelings, and environment in near real-time. While these methods can reduce recall biases typically present in survey data, they may still suffer from other biases commonly found in self-reported data (e.g., measurement error and social desirability bias). To accommodate potential biases, we develop a Bayesian hidden Markov model to simultaneously identify risk factors for subjects transitioning between discrete latent states as well as risk factors potentially associated with them misreporting their true behaviors. We use simulated data to demonstrate how ignoring potential measurement error can negatively affect variable selection performance and estimation accuracy. We apply our proposed model to smartphone-based ecological momentary assessment data collected within a randomized controlled trial that evaluated the impact of incentivizing abstinence from cigarette smoking among socioeconomically disadvantaged adults. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

A Bayesian joint model for compositional mediation effect selection in microbiome data

Analyzing multivariate count data generated by high-throughput sequencing technology in microbiome research studies is challenging due to the high-dimensional and compositional structure of the data and overdispersion. In practice, researchers are often interested in investigating how the microbiome may mediate the relation between an assigned treatment and an observed phenotypic response. Existing approaches designed for compositional mediation analysis are unable to simultaneously determine the presence of direct effects, relative indirect effects, and overall indirect effects, while quantifying their uncertainty. We propose a formulation of a Bayesian joint model for compositional data that allows for the identification, estimation, and uncertainty quantification of various causal estimands in high-dimensional mediation analysis. We conduct simulation studies and compare our method's mediation effects selection performance with existing methods. Finally, we apply our method to a benchmark data set investigating the sub-therapeutic antibiotic treatment effect on body weight in early-life mice.

Bayesian feature selection for radiomics using reliability metrics

Introduction: Imaging of tumors is a standard step in diagnosing cancer and making subsequent treatment decisions. The field of radiomics aims to develop imaging based biomarkers using methods rooted in artificial intelligence applied to medical imaging. However, a challenging aspect of developing predictive models for clinical use is that many quantitative features derived from image data exhibit instability or lack of reproducibility across different imaging systems or image-processing pipelines.

Methods: To address this challenge, we propose a Bayesian sparse modeling approach for image classification based on radiomic features, where the inclusion of more reliable features is favored via a probit prior formulation.

Results: We verify through simulation studies that this approach can improve feature selection and prediction given correct prior information. Finally, we illustrate the method with an application to the classification of head and neck cancer patients by human papillomavirus status, using as our prior information a reliability metric quantifying feature stability across different imaging systems.

BAYESIAN NON-HOMOGENEOUS HIDDEN MARKOV MODEL WITH VARIABLE SELECTION FOR INVESTIGATING DRIVERS OF SEIZURE RISK CYCLING

A major issue in the clinical management of epilepsy is the unpredictability of seizures. Yet, traditional approaches to seizure forecasting and risk assessment in epilepsy rely heavily on raw seizure frequencies, which are a stochastic measurement of seizure risk. We consider a Bayesian non-homogeneous hidden Markov model for unsupervised clustering of zero-inflated seizure count data. The proposed model allows for a probabilistic estimate of the sequence of seizure risk states at the individual level. It also offers significant improvement over prior approaches by incorporating a variable selection prior for the identification of clinical covariates that drive seizure risk changes and accommodating highly granular data. For inference, we implement an efficient sampler that employs stochastic search and data augmentation techniques. We evaluate model performance on simulated seizure count data. We then demonstrate the clinical utility of the proposed model by analyzing daily seizure count data from 133 patients with Dravet syndrome collected through the Seizure Tracker™ system, a patient-reported electronic seizure diary. We report on the dynamics of seizure risk cycling, including validation of several known pharmacologic relationships. We also uncover novel findings characterizing the presence and volatility of risk states in Dravet syndrome, which may directly inform counseling to reduce the unpredictability of seizures for patients with this devastating cause of epilepsy.

Seizure count forecasting to aid diagnostic testing in epilepsy

OBJECTIVE

Epilepsy monitoring unit (EMU) admissions are critical for presurgical evaluation of drug-resistant epilepsy but may be nondiagnostic if an insufficient number of seizures are recorded. Seizure forecasting algorithms have shown promise for estimating the likelihood of seizures as a binary event in individual patients, but methods to predict how many seizures will occur remain elusive. Such methods could increase the diagnostic yield of EMU admissions and help patients mitigate seizure-related morbidity. Here, we evaluated the performance of a state-space method that uses prior seizure count data to predict future counts.

METHODS

A Bayesian negative-binomial dynamic linear model (DLM) was developed to forecast daily electrographic seizure counts in 19 patients implanted with a responsive neurostimulation (RNS) device. Holdout validation was used to evaluate performance in predicting the number of electrographic seizures for forecast horizons ranging 1-7 days ahead.

RESULTS

One-day-ahead prediction of the number of electrographic seizures using a negative-binomial DLM resulted in improvement over chance in 73.1% of time segments compared to a random chance forecaster and remained >50% for forecast horizons of up to 7 days. Superior performance (mean error = .99) was obtained in predicting the number of electrographic seizures in the next day compared to three traditional methods for count forecasting (integer-valued generalized autoregressive conditional heteroskedasticity model or INGARCH, 1.10; Croston, 1.06; generalized linear autoregressive moving average model or GLARMA, 2.00). Number of electrographic seizures in the preceding day and laterality of electrographic pattern detections had highest predictive value, with greater number of electrographic seizures and RNS magnet swipes in the preceding day associated with a higher number of electrographic seizures the next day.

SIGNIFICANCE

This study demonstrates that DLMs can predict the number of electrographic seizures a patient will experience days in advance with above chance accuracy. This study represents an important step toward the translation of seizure forecasting methods into the optimization of EMU admissions.

Effective connectivity between resting-state networks in depression

RATIONALE

Although depression has been widely researched, findings characterizing how brain regions influence each other remains scarce, yet this is critical for research on antidepressant treatments and individual responses to particular treatments.

OBJECTIVES

To identify pre-treatment resting state effective connectivity (rsEC) patterns in patients with major depressive disorder (MDD) and explore their relationship with treatment response.

METHODS

Thirty-four drug-free MDD patients had an MRI scan and were subsequently treated for 6 weeks with an SSRI escitalopram 10 mg daily; the response was defined as ≥50% decrease in Hamilton Depression Rating Scale (HAMD) score.

RESULTS

rsEC networks in default mode, central executive, and salience networks were identified for patients with depression. Exploratory analyses indicated higher connectivity strength related to baseline depression severity and response to treatment.

CONCLUSIONS

Preliminary analyses revealed widespread dysfunction of rsEC in depression. Functional rsEC may be useful as a predictive tool for antidepressant treatment response. A primary limitation of the current study was the small size; however, the group was carefully chosen, well-characterized, and included only medication-free patients. Further research in large samples of placebo-controlled studies would be required to confirm the results.

Prevalence and features of delirium in older patients admitted to rehabilitation facilities: a multicenter study

BACKGROUND

Delirium is thought to be common across various settings of care; however, still little research has been conducted in rehabilitation.

AIM

We investigated the prevalence of delirium, its features and motor subtypes in older patients admitted to rehabilitation facilities during the three editions of the "Delirium Day project".

METHODS

We conducted a cross-sectional study in which 1237 older patients (age ≥ 65 years old) admitted to 50 Italian rehabilitation wards during the three editions of the "Delirium Day project" (2015 to 2017) were included. Delirium was evaluated through the 4AT and its motor subtype with the Delirium Motor Subtype Scale.

RESULTS

Delirium was detected in 226 patients (18%), and the most recurrent motor subtype was mixed (37%), followed by hypoactive (26%), hyperactive (21%) and non-motor one (16%). In a multivariate Poisson regression model with robust variance, factors associated with delirium were: disability in basic (PR 1.48, 95%CI: 1.17-1.9, p value 0.001) and instrumental activities of daily living (PR 1.58, 95%CI: 1.08-2.32, p value 0.018), dementia (PR 2.10, 95%CI: 1.62-2.73, p value < 0.0001), typical antipsychotics (PR 1.47, 95%CI: 1.10-1.95, p value 0.008), antidepressants other than selective serotonin reuptake inhibitors (PR 1.3, 95%CI: 1.02-1.66, p value 0.035), and physical restraints (PR 2.37, 95%CI: 1.68-3.36, p value < 0.0001).

CONCLUSION

This multicenter study reports that 2 out 10 patients admitted to rehabilitations had delirium on the index day. Mixed delirium was the most prevalent subtype. Delirium was associated with unmodifiable (dementia, disability) and modifiable (physical restraints, medications) factors. Identification of these factors should prompt specific interventions aimed to prevent or mitigate delirium.

The association between low skeletal muscle mass and delirium: results from the nationwide multi-centre Italian Delirium Day 2017

Introduction

Delirium and sarcopenia are common, although underdiagnosed, geriatric syndromes. Several pathological mechanisms can link delirium and low skeletal muscle mass, but few studies have investigated their association. We aimed to investigate (1) the association between delirium and low skeletal muscle mass and (2) the possible role of calf circumference mass in finding cases with delirium.

Methods

The analyses were conducted employing the cross-sectional “Delirium Day” initiative, on patient 65 years and older admitted to acute hospital medical wards, emergency departments, rehabilitation wards, nursing homes and hospices in Italy in 2017. Delirium was diagnosed as a 4 + score at the 4-AT scale. Low skeletal muscle mass was operationally defined as calf circumference ≤ 34 cm in males and ≤ 33 cm in females. Logistic regression models were used to investigate the association between low skeletal muscle mass and delirium. The discriminative ability of calf circumference was evaluated using non-parametric ROC analyses.

Results

A sample of 1675 patients was analyzed. In total, 73.6% of participants had low skeletal muscle mass and 24.1% exhibited delirium. Low skeletal muscle mass and delirium showed an independent association (OR: 1.50; 95% CI 1.09–2.08). In the subsample of patients without a diagnosis of dementia, the inclusion of calf circumference in a model based on age and sex significantly improved its discriminative accuracy [area under the curve (AUC) 0.69 vs 0.57, p < 0.001].

Discussion and conclusion

Low muscle mass is independently associated with delirium. In patients without a previous diagnosis of dementia, calf circumference may help to better identify those who develop delirium.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40520-021-01950-8.

Bayesian inference for stationary points in gaussian process regression models for event-related potentials analysis

Stationary points embedded in the derivatives are often critical for a model to be interpretable and may be considered as key features of interest in many applications. We propose a semiparametric Bayesian model to efficiently infer the locations of stationary points of a nonparametric function, which also produces an estimate of the function. We use Gaussian processes as a flexible prior for the underlying function and impose derivative constraints to control the function's shape via conditioning. We develop an inferential strategy that intentionally restricts estimation to the case of at least one stationary point, bypassing possible mis-specifications in the number of stationary points and avoiding the varying dimension problem that often brings in computational complexity. We illustrate the proposed methods using simulations and then apply the method to the estimation of event-related potentials (ERP) derived from electroencephalography (EEG) signals. We show how the proposed method automatically identifies characteristic components and their latencies at the individual level, which avoids the excessive averaging across subjects which is routinely done in the field to obtain smooth curves. By applying this approach to EEG data collected from younger and older adults during a speech perception task, we are able to demonstrate how the time course of speech perception processes changes with age. This article is protected by copyright. All rights reserved.

Effective connectivity in the default mode network after paediatric traumatic brain injury

Children who experience a traumatic brain injury (TBI) are at elevated risk for a range of negative cognitive and neuropsychological outcomes. Identifying which children are at greatest risk for negative outcomes can be difficult due to the heterogeneity of TBI. To address this barrier, the current study applied a novel method of characterizing brain connectivity networks, Bayesian multi-subject vector autoregressive modelling (BVAR-connect), which used white matter integrity as priors to evaluate effective connectivity-the time-dependent relationship in functional magnetic resonance imaging (fMRI) activity between two brain regions-within the default mode network (DMN). In a prospective longitudinal study, children ages 8-15 years with mild to severe TBI underwent diffusion tensor imaging and resting state fMRI 7 weeks after injury; post-concussion and anxiety symptoms were assessed 7 months after injury. The goals of this study were to (1) characterize differences in positive effective connectivity of resting-state DMN circuitry between healthy controls and children with TBI, (2) determine if severity of TBI was associated with differences in DMN connectivity and (3) evaluate whether patterns of DMN effective connectivity predicted persistent post-concussion symptoms and anxiety. Healthy controls had unique positive connectivity that mostly emerged from the inferior temporal lobes. In contrast, children with TBI had unique effective connectivity among orbitofrontal and parietal regions. These positive orbitofrontal-parietal DMN effective connectivity patterns also differed by TBI severity and were associated with persisting behavioural outcomes. Effective connectivity may be a sensitive neuroimaging marker of TBI severity as well as a predictor of chronic post-concussion symptoms and anxiety.

Latent Network Estimation and Variable Selection for Compositional Data Via Variational EM

Network estimation and variable selection have been extensively studied in the statistical literature, but only recently have those two challenges been addressed simultaneously. In this article, we seek to develop a novel method to simultaneously estimate network interactions and associations to relevant covariates for count data, and specifically for compositional data, which have a fixed sum constraint. We use a hierarchical Bayesian model with latent layers and employ spike-and-slab priors for both edge and covariate selection. For posterior inference, we develop a novel variational inference scheme with an expectation-maximization step, to enable efficient estimation. Through simulation studies, we demonstrate that the proposed model outperforms existing methods in its accuracy of network recovery. We show the practical utility of our model via an application to microbiome data. The human microbiome has been shown to contribute too many of the functions of the human body, and also to be linked with a number of diseases. In our application, we seek to better understand the interaction between microbes and relevant covariates, as well as the interaction of microbes with each other. We call our algorithm simultaneous inference for networks and covariates and provide a Python implementation, which is available online.

Evidence of state-dependence in the effectiveness of responsive neurostimulation for seizure modulation

Background:

An implanted device for brain-responsive neurostimulation (RNS^® System) is approved as an effective treatment to reduce seizures in adults with medically-refractory focal epilepsy. Clinical trials of the RNS System demonstrate population-level reduction in average seizure frequency, but therapeutic response is highly variable.

Hypothesis:

Recent evidence links seizures to cyclical fluctuations in underlying risk. We tested the hypothesis that effectiveness of responsive neurostimulation varies based on current state within cyclical risk fluctuations.

Methods:

We analyzed retrospective data from 25 adults with medically-refractory focal epilepsy implanted with the RNS System. Chronic electrocorticography was used to record electrographic seizures, and hidden Markov models decoded seizures into fluctuations in underlying risk. State-dependent associations of RNS System stimulation parameters with changes in risk were estimated.

Results:

Higher charge density was associated with improved outcomes, both for remaining in a low seizure risk state and for transitioning from a high to a low seizure risk state. The effect of stimulation frequency depended on initial seizure risk state: when starting in a low risk state, higher stimulation frequencies were associated with remaining in a low risk state, but when starting in a high risk state, lower stimulation frequencies were associated with transition to a low risk state. Findings were consistent across bipolar and monopolar stimulation configurations.

Conclusion:

The impact of RNS on seizure frequency exhibits state-dependence, such that stimulation parameters which are effective in one seizure risk state may not be effective in another. These findings represent conceptual advances in understanding the therapeutic mechanism of RNS, and directly inform current practices of RNS tuning and the development of next-generation neurostimulation systems.

Correction to: MicroBVS: Dirichlet-tree multinomial regression models with Bayesian variable selection - an R package

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

MicroBVS: Dirichlet-tree multinomial regression models with Bayesian variable selection - an R package

Background

Understanding the relation between the human microbiome and modulating factors, such as diet, may help researchers design intervention strategies that promote and maintain healthy microbial communities. Numerous analytical tools are available to help identify these relations, oftentimes via automated variable selection methods. However, available tools frequently ignore evolutionary relations among microbial taxa, potential relations between modulating factors, as well as model selection uncertainty.

Results

We present MicroBVS, an R package for Dirichlet-tree multinomial models with Bayesian variable selection, for the identification of covariates associated with microbial taxa abundance data. The underlying Bayesian model accommodates phylogenetic structure in the abundance data and various parameterizations of covariates’ prior probabilities of inclusion.

Conclusion

While developed to study the human microbiome, our software can be employed in various research applications, where the aim is to generate insights into the relations between a set of covariates and compositional data with or without a known tree-like structure.

Bayesian inference of networks across multiple sample groups and data types

In this article, we develop a graphical modeling framework for the inference of networks across multiple sample groups and data types. In medical studies, this setting arises whenever a set of subjects, which may be heterogeneous due to differing disease stage or subtype, is profiled across multiple platforms, such as metabolomics, proteomics, or transcriptomics data. Our proposed Bayesian hierarchical model first links the network structures within each platform using a Markov random field prior to relate edge selection across sample groups, and then links the network similarity parameters across platforms. This enables joint estimation in a flexible manner, as we make no assumptions on the directionality of influence across the data types or the extent of network similarity across the sample groups and platforms. In addition, our model formulation allows the number of variables and number of subjects to differ across the data types, and only requires that we have data for the same set of groups. We illustrate the proposed approach through both simulation studies and an application to gene expression levels and metabolite abundances on subjects with varying severity levels of chronic obstructive pulmonary disease. Bayesian inference; Chronic obstructive pulmonary disease (COPD); Data integration; Gaussian graphical model; Markov random field prior; Spike and slab prior.

Two-group Poisson-Dirichlet mixtures for multiple testing

The simultaneous testing of multiple hypotheses is common to the analysis of high-dimensional data sets. The two-group model, first proposed by Efron, identifies significant comparisons by allocating observations to a mixture of an empirical null and an alternative distribution. In the Bayesian nonparametrics literature, many approaches have suggested using mixtures of Dirichlet Processes in the two-group model framework. Here, we investigate employing mixtures of two-parameter Poisson-Dirichlet Processes instead, and show how they provide a more flexible and effective tool for large-scale hypothesis testing. Our model further employs nonlocal prior densities to allow separation between the two mixture components. We obtain a closed-form expression for the exchangeable partition probability function of the two-group model, which leads to a straightforward Markov Chain Monte Carlo implementation. We compare the performance of our method for large-scale inference in a simulation study and illustrate its use on both a prostate cancer data set and a case-control microbiome study of the gastrointestinal tracts in children from underdeveloped countries who have been recently diagnosed with moderate-to-severe diarrhea.

Bayesian modeling of multiple structural connectivity networks during the progression of Alzheimer's disease

Alzheimer's disease is the most common neurodegenerative disease. The aim of this study is to infer structural changes in brain connectivity resulting from disease progression using cortical thickness measurements from a cohort of participants who were either healthy control, or with mild cognitive impairment, or Alzheimer's disease patients. For this purpose, we develop a novel approach for inference of multiple networks with related edge values across groups. Specifically, we infer a Gaussian graphical model for each group within a joint framework, where we rely on Bayesian hierarchical priors to link the precision matrix entries across groups. Our proposal differs from existing approaches in that it flexibly learns which groups have the most similar edge values, and accounts for the strength of connection (rather than only edge presence or absence) when sharing information across groups. Our results identify key alterations in structural connectivity that may reflect disruptions to the healthy brain, such as decreased connectivity within the occipital lobe with increasing disease severity. We also illustrate the proposed method through simulations, where we demonstrate its performance in structure learning and precision matrix estimation with respect to alternative approaches.

Prospective validation study of an epilepsy seizure risk system for outpatient evaluation

OBJECTIVE

We conducted clinical testing of an automated Bayesian machine learning algorithm (Epilepsy Seizure Assessment Tool [EpiSAT]) for outpatient seizure risk assessment using seizure counting data, and validated performance against specialized epilepsy clinician experts.

METHODS

We conducted a prospective longitudinal study of EpiSAT performance against 24 specialized clinician experts at three tertiary referral epilepsy centers in the United States. Accuracy, interrater reliability, and intra-rater reliability of EpiSAT for correctly identifying changes in seizure risk (improvements, worsening, or no change) were evaluated using 120 seizures from four synthetic seizure diaries (seizure risk known) and 120 seizures from four real seizure diaries (seizure risk unknown). The proportion of observed agreement between EpiSAT and clinicians was evaluated to assess compatibility of EpiSAT with clinical decision patterns by epilepsy experts.

RESULTS

EpiSAT exhibited substantial observed agreement (75.4%) with clinicians for assessing seizure risk. The mean accuracy of epilepsy providers for correctly assessing seizure risk was 74.7%. EpiSAT accurately identified seizure risk in 87.5% of seizure diary entries, corresponding to a significant improvement of 17.4% (P = .002). Clinicians exhibited low-to-moderate interrater reliability for seizure risk assessment (Krippendorff's α = 0.46) with good intrarater reliability across a 4- to 12-week evaluation period (Scott's π = 0.89).

SIGNIFICANCE

These results validate the ability of EpiSAT to yield objective clinical recommendations on seizure risk which follow decision patterns similar to those from specialized epilepsy providers, but with improved accuracy and reproducibility. This algorithm may serve as a useful clinical decision support system for quantitative analysis of clinical seizure frequency in clinical epilepsy practice.

Hierarchical Normalized Completely Random Measures for Robust Graphical Modeling

Gaussian graphical models are useful tools for exploring network structures in multivariate normal data. In this paper we are interested in situations where data show departures from Gaussianity, therefore requiring alternative modeling distributions. The multivariate t-distribution, obtained by dividing each component of the data vector by a gamma random variable, is a straightforward generalization to accommodate deviations from normality such as heavy tails. Since different groups of variables may be contaminated to a different extent, Finegold and Drton (2014) introduced the Dirichlet t-distribution, where the divisors are clustered using a Dirichlet process. In this work, we consider a more general class of nonparametric distributions as the prior on the divisor terms, namely the class of normalized completely random measures (NormCRMs). To improve the effectiveness of the clustering, we propose modeling the dependence among the divisors through a nonparametric hierarchical structure, which allows for the sharing of parameters across the samples in the data set. This desirable feature enables us to cluster together different components of multivariate data in a parsimonious way. We demonstrate through simulations that this approach provides accurate graphical model inference, and apply it to a case study examining the dependence structure in radiomics data derived from The Cancer Imaging Atlas.

Individual Differences in the Neural and Cognitive Mechanisms of Single Word Reading

Written language is a human invention that our brains did not evolve for. Yet, most research has focused on finding a single theory of reading, identifying the common set of cognitive and neural processes shared across individuals, neglecting individual differences. In contrast, we investigated variation in single word reading. Using a novel statistical method for analyzing heterogeneity in multi-subject task-based functional magnetic resonance imaging (fMRI), we clustered readers based on their brain's response to written stimuli. Separate behavioral testing and neuroimaging analysis shows that these clusters differed in the role of the sublexical pathway in processing written language, but not in reading skill. Taken together, these results suggest that individuals vary in the cognitive and neural mechanisms involved in word reading. In general, neurocognitive theories need to account not only for what tends to be true of the population, but also the types of variation that exist, even within a neurotypical population.

A Bayesian Approach for Estimating Dynamic Functional Network Connectivity in fMRI Data

Dynamic functional connectivity, i.e., the study of how interactions among brain regions change dynamically over the course of an fMRI experiment, has recently received wide interest in the neuroimaging literature. Current approaches for studying dynamic connectivity often rely on ad-hoc approaches for inference, with the fMRI time courses segmented by a sequence of sliding windows. We propose a principled Bayesian approach to dynamic functional connectivity, which is based on the estimation of time varying networks. Our method utilizes a hidden Markov model for classification of latent cognitive states, achieving estimation of the networks in an integrated framework that borrows strength over the entire time course of the experiment. Furthermore, we assume that the graph structures, which define the connectivity states at each time point, are related within a super-graph, to encourage the selection of the same edges among related graphs. We apply our method to simulated task-based fMRI data, where we show how our approach allows the decoupling of the task-related activations and the functional connectivity states. We also analyze data from an fMRI sensorimotor task experiment on an individual healthy subject and obtain results that support the role of particular anatomical regions in modulating interaction between executive control and attention networks.

Epilepsy as a dynamic disease: A Bayesian model for differentiating seizure risk from natural variability

Objective

A fundamental challenge in treating epilepsy is that changes in observed seizure frequencies do not necessarily reflect changes in underlying seizure risk. Rather, changes in seizure frequency may occur due to probabilistic variation around an underlying seizure risk state caused by normal fluctuations from natural history, leading to seizure unpredictability and potentially suboptimal medication adjustments in epilepsy management. However, no rigorous statistical approach exists to systematically distinguish expected changes in seizure frequency due to natural variability from changes in underlying seizure risk.

Methods

Using data from SeizureTracker.com, a patient‐reported seizure diary tool containing over 1.2 million recorded seizures across 8 years, a novel epilepsy seizure risk assessment tool (EpiSAT) employing a Bayesian mixed‐effects hidden Markov model for zero‐inflated count data was developed to estimate changes in underlying seizure risk using patient‐reported seizure diary and clinical measurement data. Accuracy for correctly assessing underlying seizure risk was evaluated through a simulation comparison. Implications for the natural history of tuberous sclerosis complex (TSC) were assessed using data from SeizureTracker.com.

Results

EpiSAT led to significant improvement in seizure risk assessment compared to traditional approaches relying solely on observed seizure frequencies. Applied to TSC, four underlying seizure risk states were identified. The expected duration of each state was <12 months, providing a data‐driven estimate of the amount of time a person with TSC would be expected to remain at the same seizure risk level according to the natural course of epilepsy.

Significance

We propose a novel Bayesian statistical approach for evaluating seizure risk on an individual patient level using patient‐reported seizure diaries, which allows for the incorporation of external clinical variables to assess impact on seizure risk. This tool may improve the ability to distinguish true changes in seizure risk from natural variations in seizure frequency in clinical practice. Incorporation of systematic statistical approaches into antiepileptic drug (AED) management may help improve understanding of seizure unpredictability as well as timing of treatment interventions for people with epilepsy.

Transcriptomic analysis reveals inflammatory and metabolic pathways that are regulated by renal perfusion pressure in the outer medulla of Dahl-S rats

Studies exploring the development of hypertension have traditionally been unable to distinguish which of the observed changes are underlying causes from those that are a consequence of elevated blood pressure. In this study, a custom-designed servo-control system was utilized to precisely control renal perfusion pressure to the left kidney continuously during the development of hypertension in Dahl salt-sensitive rats. In this way, we maintained the left kidney at control blood pressure while the right kidney was exposed to hypertensive pressures. As each kidney was exposed to the same circulating factors, differences between them represent changes induced by pressure alone. RNA sequencing analysis identified 1,613 differently expressed genes affected by renal perfusion pressure. Three pathway analysis methods were applied, one a novel approach incorporating arterial pressure as an input variable allowing a more direct connection between the expression of genes and pressure. The statistical analysis proposed several novel pathways by which pressure affects renal physiology. We confirmed the effects of pressure on p-Jnk regulation, in which the hypertensive medullas show increased p-Jnk/Jnk ratios relative to the left (0.79 ± 0.11 vs. 0.53 ± 0.10, P < 0.01, n = 8). We also confirmed pathway predictions of mitochondrial function, in which the respiratory control ratio of hypertensive vs. control mitochondria are significantly reduced (7.9 ± 1.2 vs. 10.4 ± 1.8, P < 0.01, n = 6) and metabolomic profile, in which 14 metabolites differed significantly between hypertensive and control medullas ( P < 0.05, n = 5). These findings demonstrate that subtle differences in the transcriptome can be used to predict functional changes of the kidney as a consequence of pressure elevation.

Temporal and spectral characteristics of dynamic functional connectivity between resting-state networks reveal information beyond static connectivity

Estimation of functional connectivity (FC) has become an increasingly powerful tool for investigating healthy and abnormal brain function. Static connectivity, in particular, has played a large part in guiding conclusions from the majority of resting-state functional MRI studies. However, accumulating evidence points to the presence of temporal fluctuations in FC, leading to increasing interest in estimating FC as a dynamic quantity. One central issue that has arisen in this new view of connectivity is the dramatic increase in complexity caused by dynamic functional connectivity (dFC) estimation. To computationally handle this increased complexity, a limited set of dFC properties, primarily the mean and variance, have generally been considered. Additionally, it remains unclear how to integrate the increased information from dFC into pattern recognition techniques for subject-level prediction. In this study, we propose an approach to address these two issues based on a large number of previously unexplored temporal and spectral features of dynamic functional connectivity. A Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model is used to estimate time-varying patterns of functional connectivity between resting-state networks. Time-frequency analysis is then performed on dFC estimates, and a large number of previously unexplored temporal and spectral features drawn from signal processing literature are extracted for dFC estimates. We apply the investigated features to two neurologic populations of interest, healthy controls and patients with temporal lobe epilepsy, and show that the proposed approach leads to substantial increases in predictive performance compared to both traditional estimates of static connectivity as well as current approaches to dFC. Variable importance is assessed and shows that there are several quantities that can be extracted from dFC signal which are more informative than the traditional mean or variance of dFC. This work illuminates many previously unexplored facets of the dynamic properties of functional connectivity between resting-state networks, and provides a platform for dynamic functional connectivity analysis that facilitates its usage as an investigative measure for healthy as well as abnormal brain function.

A Bayesian Approach for Learning Gene Networks Underlying Disease Severity in COPD

In this paper, we propose a Bayesian hierarchical approach to infer network structures across multiple sample groups where both shared and differential edges may exist across the groups. In our approach, we link graphs through a Markov random field prior. This prior on network similarity provides a measure of pairwise relatedness that borrows strength only between related groups. We incorporate the computational efficiency of continuous shrinkage priors, improving scalability for network estimation in cases of larger dimensionality. Our model is applied to patient groups with increasing levels of chronic obstructive pulmonary disease severity, with the goal of better understanding the break down of gene pathways as the disease progresses. Our approach is able to identify critical hub genes for four targeted pathways. Furthermore, it identifies gene connections that are disrupted with increased disease severity and that characterize the disease evolution. We also demonstrate the superior performance of our approach with respect to competing methods, using simulated data.

A fully Bayesian latent variable model for integrative clustering analysis of multi-type omics data

Identification of clinically relevant tumor subtypes and omics signatures is an important task in cancer translational research for precision medicine. Large-scale genomic profiling studies such as The Cancer Genome Atlas (TCGA) Research Network have generated vast amounts of genomic, transcriptomic, epigenomic, and proteomic data. While these studies have provided great resources for researchers to discover clinically relevant tumor subtypes and driver molecular alterations, there are few computationally efficient methods and tools for integrative clustering analysis of these multi-type omics data. Therefore, the aim of this article is to develop a fully Bayesian latent variable method (called iClusterBayes) that can jointly model omics data of continuous and discrete data types for identification of tumor subtypes and relevant omics features. Specifically, the proposed method uses a few latent variables to capture the inherent structure of multiple omics data sets to achieve joint dimension reduction. As a result, the tumor samples can be clustered in the latent variable space and relevant omics features that drive the sample clustering are identified through Bayesian variable selection. This method significantly improve on the existing integrative clustering method iClusterPlus in terms of statistical inference and computational speed. By analyzing TCGA and simulated data sets, we demonstrate the excellent performance of the proposed method in revealing clinically meaningful tumor subtypes and driver omics features.

A Hierarchical Bayesian Model for the Identification of PET Markers Associated to the Prediction of Surgical Outcome after Anterior Temporal Lobe Resection

We develop an integrative Bayesian predictive modeling framework that identifies individual pathological brain states based on the selection of fluoro-deoxyglucose positron emission tomography (PET) imaging biomarkers and evaluates the association of those states with a clinical outcome. We consider data from a study on temporal lobe epilepsy (TLE) patients who subsequently underwent anterior temporal lobe resection. Our modeling framework looks at the observed profiles of regional glucose metabolism in PET as the phenotypic manifestation of a latent individual pathologic state, which is assumed to vary across the population. The modeling strategy we adopt allows the identification of patient subgroups characterized by latent pathologies differentially associated to the clinical outcome of interest. It also identifies imaging biomarkers characterizing the pathological states of the subjects. In the data application, we identify a subgroup of TLE patients at high risk for post-surgical seizure recurrence after anterior temporal lobe resection, together with a set of discriminatory brain regions that can be used to distinguish the latent subgroups. We show that the proposed method achieves high cross-validated accuracy in predicting post-surgical seizure recurrence.

Identifying Biomarkers from Mass Spectrometry Data with Ordinal Outcome

In recent years, there has been an increased interest in using protein mass spectroscopy to identify molecular markers that discriminate diseased from healthy individuals. Existing methods are tailored towards classifying observations into nominal categories. Sometimes, however, the outcome of interest may be measured on an ordered scale. Ignoring this natural ordering results in some loss of information. In this paper, we propose a Bayesian model for the analysis of mass spectrometry data with ordered outcome. The method provides a unified approach for identifying relevant markers and predicting class membership. This is accomplished by building a stochastic search variable selection method within an ordinal outcome model. We apply the methodology to mass spectrometry data on ovarian cancer cases and healthy individuals. We also utilize wavelet-based techniques to remove noise from the mass spectra prior to analysis. We identify protein markers associated with being healthy, having low grade ovarian cancer, or being a high grade case. For comparison, we repeated the analysis using conventional classification procedures and found improved predictive accuracy with our method.

Bayesian variable selection for a semi-competing risks model with three hazard functions

A variable selection procedure is developed for a semi-competing risks regression model with three hazard functions that uses spike-and-slab priors and stochastic search variable selection algorithms for posterior inference. A rule is devised for choosing the threshold on the marginal posterior probability of variable inclusion based on the Deviance Information Criterion (DIC) that is examined in a simulation study. The method is applied to data from esophageal cancer patients from the MD Anderson Cancer Center, Houston, TX, where the most important covariates are selected in each of the hazards of effusion, death before effusion, and death after effusion. The DIC procedure that is proposed leads to similar selected models regardless of the choices of some of the hyperparameters. The application results show that patients with intensity-modulated radiation therapy have significantly reduced risks of pericardial effusion, pleural effusion, and death before either effusion type.

An integrative Bayesian Dirichlet-multinomial regression model for the analysis of taxonomic abundances in microbiome data

Background

The Human Microbiome has been variously associated with the immune-regulatory mechanisms involved in the prevention or development of many non-infectious human diseases such as autoimmunity, allergy and cancer. Integrative approaches which aim at associating the composition of the human microbiome with other available information, such as clinical covariates and environmental predictors, are paramount to develop a more complete understanding of the role of microbiome in disease development.

Results

In this manuscript, we propose a Bayesian Dirichlet-Multinomial regression model which uses spike-and-slab priors for the selection of significant associations between a set of available covariates and taxa from a microbiome abundance table. The approach allows straightforward incorporation of the covariates through a log-linear regression parametrization of the parameters of the Dirichlet-Multinomial likelihood. Inference is conducted through a Markov Chain Monte Carlo algorithm, and selection of the significant covariates is based upon the assessment of posterior probabilities of inclusions and the thresholding of the Bayesian false discovery rate. We design a simulation study to evaluate the performance of the proposed method, and then apply our model on a publicly available dataset obtained from the Human Microbiome Project which associates taxa abundances with KEGG orthology pathways. The method is implemented in specifically developed R code, which has been made publicly available.

Conclusions

Our method compares favorably in simulations to several recently proposed approaches for similarly structured data, in terms of increased accuracy and reduced false positive as well as false negative rates. In the application to the data from the Human Microbiome Project, a close evaluation of the biological significance of our findings confirms existing associations in the literature.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-017-1516-0) contains supplementary material, which is available to authorized users.

Bayesian vector autoregressive model for multi-subject effective connectivity inference using multi-modal neuroimaging data

In this article a multi-subject vector autoregressive (VAR) modeling approach was proposed for inference on effective connectivity based on resting-state functional MRI data. Their framework uses a Bayesian variable selection approach to allow for simultaneous inference on effective connectivity at both the subject- and group-level. Furthermore, it accounts for multi-modal data by integrating structural imaging information into the prior model, encouraging effective connectivity between structurally connected regions. They demonstrated through simulation studies that their approach resulted in improved inference on effective connectivity at both the subject- and group-level, compared with currently used methods. It was concluded by illustrating the method on temporal lobe epilepsy data, where resting-state functional MRI and structural MRI were used. Hum Brain Mapp 38:1311-1332, 2017. © 2016 Wiley Periodicals, Inc.

KScons: a Bayesian approach for protein residue contact prediction using the knob-socket model of protein tertiary structure

MOTIVATION

By simplifying the many-bodied complexity of residue packing into patterns of simple pairwise secondary structure interactions between a single knob residue with a three-residue socket, the knob-socket construct allows a more direct incorporation of structural information into the prediction of residue contacts. By modeling the preferences between the amino acid composition of a socket and knob, we undertake an investigation of the knob-socket construct's ability to improve the prediction of residue contacts. The statistical model considers three priors and two posterior estimations to better understand how the input data affects predictions. This produces six implementations of KScons that are tested on three sets: PSICOV, CASP10 and CASP11. We compare against the current leading contact prediction methods.

RESULTS

The results demonstrate the usefulness as well as the limits of knob-socket based structural modeling of protein contacts. The construct is able to extract good predictions from known structural homologs, while its performance degrades when no homologs exist. Among our six implementations, KScons MST-MP (which uses the multiple structure alignment prior and marginal posterior incorporating structural homolog information) performs the best in all three prediction sets. An analysis of recall and precision finds that KScons MST-MP improves accuracy not only by improving identification of true positives, but also by decreasing the number of false positives. Over the CASP10 and CASP11 sets, KScons MST-MP performs better than the leading methods using only evolutionary coupling data, but not quite as well as the supervised learning methods of MetaPSICOV and CoinDCA-NN that incorporate a large set of structural features.

CONTACT

qiwei.li@rice.eduSupplementary information: Supplementary data are available at Bioinformatics online.

A Bayesian Nonparametric Approach for Functional Data Classification with Application to Hepatic Tissue Characterization

Computed tomography perfusion (CTp) is an emerging functional imaging technology that provides a quantitative assessment of the passage of fluid through blood vessels. Tissue perfusion plays a critical role in oncology due to the proliferation of networks of new blood vessels typical of cancer angiogenesis, which triggers modifications to the vasculature of the surrounding host tissue. In this article, we consider a Bayesian semiparametric model for the analysis of functional data. This method is applied to a study of four interdependent hepatic perfusion CT characteristics that were acquired under the administration of contrast using a sequence of repeated scans over a period of 590 seconds. More specifically, our modeling framework facilitates borrowing of information across patients and tissues. Additionally, the approach enables flexible estimation of temporal correlation structures exhibited by mappings of the correlated perfusion biomarkers and thus accounts for the heteroskedasticity typically observed in those measurements, by incorporating change-points in the covariance estimation. This method is applied to measurements obtained from regions of liver surrounding malignant and benign tissues, for each perfusion biomarker. We demonstrate how to cluster the liver regions on the basis of their CTp profiles, which can be used in a prediction context to classify regions of interest provided by future patients, and thereby assist in discriminating malignant from healthy tissue regions in diagnostic settings.

A Network Biology Approach Identifies Molecular Cross-Talk between Normal Prostate Epithelial and Prostate Carcinoma Cells

The advent of functional genomics has enabled the genome-wide characterization of the molecular state of cells and tissues, virtually at every level of biological organization. The difficulty in organizing and mining this unprecedented amount of information has stimulated the development of computational methods designed to infer the underlying structure of regulatory networks from observational data. These important developments had a profound impact in biological sciences since they triggered the development of a novel data-driven investigative approach. In cancer research, this strategy has been particularly successful. It has contributed to the identification of novel biomarkers, to a better characterization of disease heterogeneity and to a more in depth understanding of cancer pathophysiology. However, so far these approaches have not explicitly addressed the challenge of identifying networks representing the interaction of different cell types in a complex tissue. Since these interactions represent an essential part of the biology of both diseased and healthy tissues, it is of paramount importance that this challenge is addressed. Here we report the definition of a network reverse engineering strategy designed to infer directional signals linking adjacent cell types within a complex tissue. The application of this inference strategy to prostate cancer genome-wide expression profiling data validated the approach and revealed that normal epithelial cells exert an anti-tumour activity on prostate carcinoma cells. Moreover, by using a Bayesian hierarchical model integrating genetics and gene expression data and combining this with survival analysis, we show that the expression of putative cell communication genes related to focal adhesion and secretion is affected by epistatic gene copy number variation and it is predictive of patient survival. Ultimately, this study represents a generalizable approach to the challenge of deciphering cell communication networks in a wide spectrum of biological systems.

In the current era of cancer research, stimulated by the release of the entire human genome, it has become increasingly clear that to understand cancer we need to understand how the many thousands of genes and proteins involved interact. Modern techniques have enabled the collection of unprecedented amounts of high quality data describing the state of these molecules during cancer development. In cancer research particularly, this strategy has been particularly successful, leading to the discovery of new drugs able to target key factors promoting cancer growth. However, a large body of research suggests that in complex organs, the interaction between cancer and its surrounding environment is an essential part of the biology of both diseased and healthy tissues, therefore it is of paramount importance that this process is further investigated. Here we report a strategy designed to reveal communication signals between cancer cells and adjacent cell types. We apply the strategy to prostate cancer and find that normal cells surrounding the tumour do exert an anti-tumour activity on prostate cancer cells. By using a statistical model which integrates multiple levels of genetic data, we show that cell-to-cell communication genes are controlled by DNA alterations and have potential prognostic value.

Salient body image concerns of patients with cancer undergoing head and neck reconstruction

BACKGROUND

Patients with cancer undergoing head and neck reconstruction can experience significant distress from alterations in appearance and bodily functioning. We sought to delineate salient dimensions of body image concerns in this patient population preparing for reconstructive surgery.

METHODS

Participants completed self-report questionnaires evaluating numerous aspects of body image. We used Bayesian factor analysis modeling methods to identify latent factors emerging from the data.

RESULTS

We identified 2 latent factors: appearance distress and functional difficulties. The highest level of preoperative body image concerns were related to distress about appearance changes and its perceived social consequences. Appearance distress items displayed greater variability compared with functional difficulties.

CONCLUSION

Appearance and functional changes to body image are important areas of concern for patients with head and neck cancer as they prepare for reconstructive surgery. Knowledge regarding specific body image issues can be used to guide psychosocial assessments and intervention to enhance patient care. © 2016 Wiley Periodicals, Inc. Head Neck 38: 1035-1042, 2016.

Joint Bayesian variable and graph selection for regression models with network-structured predictors

In this work, we develop a Bayesian approach to perform selection of predictors that are linked within a network. We achieve this by combining a sparse regression model relating the predictors to a response variable with a graphical model describing conditional dependencies among the predictors. The proposed method is well-suited for genomic applications because it allows the identification of pathways of functionally related genes or proteins that impact an outcome of interest. In contrast to previous approaches for network-guided variable selection, we infer the network among predictors using a Gaussian graphical model and do not assume that network information is available a priori. We demonstrate that our method outperforms existing methods in identifying network-structured predictors in simulation settings and illustrate our proposed model with an application to inference of proteins relevant to glioblastoma survival.

Time-dependence of graph theory metrics in functional connectivity analysis

Brain graphs provide a useful way to computationally model the network structure of the connectome, and this has led to increasing interest in the use of graph theory to quantitate and investigate the topological characteristics of the healthy brain and brain disorders on the network level. The majority of graph theory investigations of functional connectivity have relied on the assumption of temporal stationarity. However, recent evidence increasingly suggests that functional connectivity fluctuates over the length of the scan. In this study, we investigate the stationarity of brain network topology using a Bayesian hidden Markov model (HMM) approach that estimates the dynamic structure of graph theoretical measures of whole-brain functional connectivity. In addition to extracting the stationary distribution and transition probabilities of commonly employed graph theory measures, we propose two estimators of temporal stationarity: the S-index and N-index. These indexes can be used to quantify different aspects of the temporal stationarity of graph theory measures. We apply the method and proposed estimators to resting-state functional MRI data from healthy controls and patients with temporal lobe epilepsy. Our analysis shows that several graph theory measures, including small-world index, global integration measures, and betweenness centrality, may exhibit greater stationarity over time and therefore be more robust. Additionally, we demonstrate that accounting for subject-level differences in the level of temporal stationarity of network topology may increase discriminatory power in discriminating between disease states. Our results confirm and extend findings from other studies regarding the dynamic nature of functional connectivity, and suggest that using statistical models which explicitly account for the dynamic nature of functional connectivity in graph theory analyses may improve the sensitivity of investigations and consistency across investigations.

Bayesian graphical network analyses reveal complex biological interactions specific to Alzheimer's disease

With different approaches to finding prognostic or diagnostic biomarkers for Alzheimer's disease (AD), many studies pursue only brief lists of biomarkers or disease specific pathways, potentially dismissing information from groups of correlated biomarkers. Using a novel Bayesian graphical network method, with data from the Australian Imaging, Biomarkers and Lifestyle (AIBL) study of aging, the aim of this study was to assess the biological connectivity between AD associated blood-based proteins. Briefly, three groups of protein markers (18, 37, and 48 proteins, respectively) were assessed for the posterior probability of biological connection both within and between clinical classifications. Clinical classification was defined in four groups: high performance healthy controls (hpHC), healthy controls (HC), participants with mild cognitive impairment (MCI), and participants with AD. Using the smaller group of proteins, posterior probabilities of network similarity between clinical classifications were very high, indicating no difference in biological connections between groups. Increasing the number of proteins increased the capacity to separate both hpHC and HC apart from the AD group (0 for complete separation, 1 for complete similarity), with posterior probabilities shifting from 0.89 for the 18 protein group, through to 0.54 for the 37 protein group, and finally 0.28 for the 48 protein group. Using this approach, we identified beta-2 microglobulin (β2M) as a potential master regulator of multiple proteins across all classifications, demonstrating that this approach can be used across many data sets to identify novel insights into diseases like AD.

Bayesian Inference of Multiple Gaussian Graphical Models

In this paper, we propose a Bayesian approach to inference on multiple Gaussian graphical models. Specifically, we address the problem of inferring multiple undirected networks in situations where some of the networks may be unrelated, while others share common features. We link the estimation of the graph structures via a Markov random field (MRF) prior which encourages common edges. We learn which sample groups have a shared graph structure by placing a spike-and-slab prior on the parameters that measure network relatedness. This approach allows us to share information between sample groups, when appropriate, as well as to obtain a measure of relative network similarity across groups. Our modeling framework incorporates relevant prior knowledge through an edge-specific informative prior and can encourage similarity to an established network. Through simulations, we demonstrate the utility of our method in summarizing relative network similarity and compare its performance against related methods. We find improved accuracy of network estimation, particularly when the sample sizes within each subgroup are moderate. We also illustrate the application of our model to infer protein networks for various cancer subtypes and under different experimental conditions.