Genomic loci influence patterns of structural covariance in the human brain.

Normal and pathologic neurobiological processes influence brain morphology in coordinated ways that give rise to patterns of structural covariance (PSC) across brain regions and individuals during brain aging and diseases. The genetic underpinnings of these patterns remain largely unknown. We apply a stochastic multivariate factorization method to a diverse population of 50,699 individuals (12 studies and 130 sites) and derive data-driven, multi-scale PSCs of regional brain size. PSCs were significantly correlated with 915 genomic loci in the discovery set, 617 of which are newly identified, and 72% were independently replicated. Key pathways influencing PSCs involve reelin signaling, apoptosis, neurogenesis, and appendage development, while pathways of breast cancer indicate potential interplays between brain metastasis and PSCs associated with neurodegeneration and dementia. Using support vector machines, multi-scale PSCs effectively derive imaging signatures of several brain diseases. Our results elucidate genetic and biological underpinnings that influence structural covariance patterns in the human brain.

Integrative analysis of multi-omics and imaging data with incorporation of biological information via structural Bayesian factor analysis.

MOTIVATION: With the rapid development of modern technologies, massive data are available for the systematic study of Alzheimer's disease (AD). Though many existing AD studies mainly focus on single-modality omics data, multi-omics datasets can provide a more comprehensive understanding of AD. To bridge this gap, we proposed a novel structural Bayesian factor analysis framework (SBFA) to extract the information shared by multi-omics data through the aggregation of genotyping data, gene expression data, neuroimaging phenotypes and prior biological network knowledge. Our approach can extract common information shared by different modalities and encourage biologically related features to be selected, guiding future AD research in a biologically meaningful way. METHOD: Our SBFA model decomposes the mean parameters of the data into a sparse factor loading matrix and a factor matrix, where the factor matrix represents the common information extracted from multi-omics and imaging data. Our framework is designed to incorporate prior biological network information. Our simulation study demonstrated that our proposed SBFA framework could achieve the best performance compared with the other state-of-the-art factor-analysis-based integrative analysis methods. RESULTS: We apply our proposed SBFA model together with several state-of-the-art factor analysis models to extract the latent common information from genotyping, gene expression and brain imaging data simultaneously from the ADNI biobank database. The latent information is then used to predict the functional activities questionnaire score, an important measurement for diagnosis of AD quantifying subjects' abilities in daily life. Our SBFA model shows the best prediction performance compared with the other factor analysis models. AVAILABILITY: Code are publicly available at https://github.com/JingxuanBao/SBFA. CONTACT: qlong@upenn.edu.

Deep learning-based identification of genetic variants: application to Alzheimer's disease classification.

Deep learning is a promising tool that uses nonlinear transformations to extract features from high-dimensional data. Deep learning is challenging in genome-wide association studies (GWAS) with high-dimensional genomic data. Here we propose a novel three-step approach (SWAT-CNN) for identification of genetic variants using deep learning to identify phenotype-related single nucleotide polymorphisms (SNPs) that can be applied to develop accurate disease classification models. In the first step, we divided the whole genome into nonoverlapping fragments of an optimal size and then ran convolutional neural network (CNN) on each fragment to select phenotype-associated fragments. In the second step, using a Sliding Window Association Test (SWAT), we ran CNN on the selected fragments to calculate phenotype influence scores (PIS) and identify phenotype-associated SNPs based on PIS. In the third step, we ran CNN on all identified SNPs to develop a classification model. We tested our approach using GWAS data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) including (N = 981; cognitively normal older adults (CN) = 650 and AD = 331). Our approach identified the well-known APOE region as the most significant genetic locus for AD. Our classification model achieved an area under the curve (AUC) of 0.82, which was compatible with traditional machine learning approaches, random forest and XGBoost. SWAT-CNN, a novel deep learning-based genome-wide approach, identified AD-associated SNPs and a classification model for AD and may hold promise for a range of biomedical applications.

Paravascular fluid dynamics reveal arterial stiffness assessed using dynamic diffusion-weighted imaging.

Paravascular cerebrospinal fluid (pCSF) surrounding the cerebral arteries within the glymphatic system is pulsatile and moves in synchrony with the pressure waves of the vessel wall. Whether such pulsatile pCSF can infer pulse wave propagation-a property tightly related to arterial stiffness-is unknown and has never been explored. Our recently developed imaging technique, dynamic diffusion-weighted imaging (dynDWI), captures the pulsatile pCSF dynamics in vivo and can explore this question. In this work, we evaluated the time shifts between pCSF waves and finger pulse waves, where pCSF waves were measured by dynDWI and finger pulse waves were measured by the scanner's built-in finger pulse oximeter. We hypothesized that the time shifts reflect brain-finger pulse wave travel time and are sensitive to arterial stiffness. We applied the framework to 36 participants aged 18-82 years to study the age effect of travel time, as well as its associations with cognitive function within the older participants (N = 15, age > 60 years). Our results revealed a strong and consistent correlation between pCSF pulse and finger pulse (mean CorrCoeff = 0.66), supporting arterial pulsation as a major driver for pCSF dynamics. The time delay between pCSF and finger pulses (TimeDelay) was significantly lower (i.e., faster pulse propagation) with advanced age (Pearson's r = -0.44, p = 0.007). Shorter TimeDelay was further associated with worse cognitive function in the older participants. Overall, our study demonstrated pCSF as a viable pathway for measuring intracranial pulses and encouraged future studies to investigate its relevance with cerebrovascular functions.

Relationship of Hoarding and Depression Symptoms in Older Adults.

Hoarding disorder (HD) is a debilitating neuropsychiatric condition that affects 2%-6% of the population and increases in incidence with age. Major depressive disorder (MDD) co-occurs with HD in approximately 50% of cases and leads to increased functional impairment and disability. However, only one study to date has examined the rate and trajectory of hoarding symptoms in older individuals with a lifetime history of MDD, including those with current active depression (late-life depression; LLD). We therefore sought to characterize this potentially distinct phenotype. We determined the incidence of HD in two separate cohorts of participants with LLD (n = 73) or lifetime history of MDD (n = 580) and examined the reliability and stability of hoarding symptoms using the Saving Inventory-Revised (SI-R) and Hoarding Rating Scale-Self Report (HRS), as well as the co-variance of hoarding and depression scores over time. HD was present in 12% to 33% of participants with MDD, with higher rates found in those with active depressive symptoms. Hoarding severity was stable across timepoints in both samples (all correlations >0.75), and fewer than 30% of participants in each sample experienced significant changes in severity between any two timepoints. Change in depression symptoms over time did not co-vary with change in hoarding symptoms. These findings indicate that hoarding is a more common comorbidity in LLD than previously suggested, and should be considered in screening and management of LLD. Future studies should further characterize the interaction of these conditions and their impact on outcomes, particularly functional impairment in this vulnerable population.

Preference matrix guided sparse canonical correlation analysis for mining brain imaging genetic associations in Alzheimer's disease.

Investigating the relationship between genetic variation and phenotypic traits is a key issue in quantitative genetics. Specifically for Alzheimer's disease, the association between genetic markers and quantitative traits remains vague while, once identified, will provide valuable guidance for the study and development of genetics-based treatment approaches. Currently, to analyze the association of two modalities, sparse canonical correlation analysis (SCCA) is commonly used to compute one sparse linear combination of the variable features for each modality, giving a pair of linear combination vectors in total that maximizes the cross-correlation between the analyzed modalities. One drawback of the plain SCCA model is that the existing findings and knowledge cannot be integrated into the model as priors to help extract interesting correlations as well as identify biologically meaningful genetic and phenotypic markers. To bridge this gap, we introduce preference matrix guided SCCA (PM-SCCA) that not only takes priors encoded as a preference matrix but also maintains computational simplicity. A simulation study and a real-data experiment are conducted to investigate the effectiveness of the model. Both experiments demonstrate that the proposed PM-SCCA model can capture not only genotype-phenotype correlation but also relevant features effectively.

Anxiety in late-life depression: Associations with brain volume, amyloid beta, white matter lesions, cognition, and functional ability.

OBJECTIVES: Late-life depression (LLD) is common and frequently co-occurs with neurodegenerative diseases of aging. Little is known about how heterogeneity within LLD relates to factors typically associated with neurodegeneration. Varying levels of anxiety are one source of heterogeneity in LLD. We examined associations between anxiety symptom severity and factors associated with neurodegeneration, including regional brain volumes, amyloid beta (Aß) deposition, white matter disease, cognitive dysfunction, and functional ability in LLD. PARTICIPANTS AND MEASUREMENTS: Older adults with major depression (N = 121, Ages 65-91) were evaluated for anxiety severity and the following: brain volume (orbitofrontal cortex [OFC], insula), cortical Aß standardized uptake value ratio (SUVR), white matter hyperintensity (WMH) volume, global cognition, and functional ability. Separate linear regression analyses adjusting for age, sex, and concurrent depression severity were conducted to examine associations between anxiety and each of these factors. A global regression analysis was then conducted to examine the relative associations of these variables with anxiety severity. RESULTS: Greater anxiety severity was associated with lower OFC volume (ß = -68.25, t = -2.18, p = .031) and greater cognitive dysfunction (ß = 0.23, t = 2.46, p = .016). Anxiety severity was not associated with insula volume, Aß SUVR, WMH, or functional ability. When examining the relative associations of cognitive functioning and OFC volume with anxiety in a global model, cognitive dysfunction (ß = 0.24, t = 2.62, p = .010), but not OFC volume, remained significantly associated with anxiety. CONCLUSIONS: Among multiple factors typically associated with neurodegeneration, cognitive dysfunction stands out as a key factor associated with anxiety severity in LLD which has implications for cognitive and psychiatric interventions.

Transcriptional risk scores in Alzheimer's disease: From pathology to cognition.

INTRODUCTION: Our previously developed blood-based transcriptional risk scores (TRS) showed associations with diagnosis and neuroimaging biomarkers for Alzheimer's disease (AD). Here, we developed brain-based TRS. METHODS: We integrated AD genome-wide association study summary and expression quantitative trait locus data to prioritize target genes using Mendelian randomization. We calculated TRS using brain transcriptome data of two independent cohorts (N = 878) and performed association analysis of TRS with diagnosis, amyloidopathy, tauopathy, and cognition. We compared AD classification performance of TRS with polygenic risk scores (PRS). RESULTS: Higher TRS values were significantly associated with AD, amyloidopathy, tauopathy, worse cognition, and faster cognitive decline, which were replicated in an independent cohort. The AD classification performance of PRS was increased with the inclusion of TRS up to 16% with the area under the curve value of 0.850. DISCUSSION: Our results suggest brain-based TRS improves the AD classification of PRS and may be a potential AD biomarker. HIGHLIGHTS: Transcriptional risk score (TRS) is developed using brain RNA-Seq data. Higher TRS values are shown in Alzheimer's disease (AD). TRS improves the AD classification power of PRS up to 16%. TRS is associated with AD pathology presence. TRS is associated with worse cognitive performance and faster cognitive decline.

Associations Between Social Network Characteristics and Brain Structure Among Older Adults.

INTRODUCTION: Social connectedness is associated with slower cognitive decline among older adults. Recent research suggests that distinct aspects of social networks may have differential effects on cognitive resilience, but few studies analyze brain structure. METHODS: This study includes 117 cognitively impaired and 59 unimpaired older adults. The effects of social network characteristics (bridging/bonding) on brain regions of interests were analyzed using linear regressions and voxel-wise multiple linear regressions of gray matter density. RESULTS: Increased social bridging was associated with greater bilateral amygdala volume and insular thickness, and left frontal lobe thickness, putamen, and thalamic volumes. Increased social bonding was associated with greater bilateral medial orbitofrontal and caudal anterior cingulate thickness, as well as right frontal lobe thickness, putamen, and amygdala volumes. DISCUSSION: The associations between social connectedness and brain structure vary depending on the types of social enrichment accessible through social networks, suggesting that psychosocial interventions could mitigate neurodegeneration. HIGHLIGHTS: Distinct forms of social capital are uniquely linked to gray matter density (GMD). Bridging is associated with preserved GMD in limbic system structures. Bonding is associated with preserved GMD in frontal lobe regions. Bridging is associated with increased brain reserve in sensory processing regions. Bonding is associated with increased brain reserve in regions of stress modulation.

Cognitively defined Alzheimer's dementia subgroups have distinct atrophy patterns.

INTRODUCTION: We sought to determine structural magnetic resonance imaging (MRI) characteristics across subgroups defined based on relative cognitive domain impairments using data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) and to compare cognitively defined to imaging-defined subgroups. METHODS: We used data from 584 people with Alzheimer's disease (AD) (461 amyloid positive, 123 unknown amyloid status) and 118 amyloid-negative controls. We used voxel-based morphometry to compare gray matter volume (GMV) for each group compared to controls and to AD-Memory. RESULTS: There was pronounced bilateral lower medial temporal lobe atrophy with relative cortical sparing for AD-Memory, lower left hemisphere GMV for AD-Language, anterior lower GMV for AD-Executive, and posterior lower GMV for AD-Visuospatial. Formal asymmetry comparisons showed substantially more asymmetry in the AD-Language group than any other group (p = 1.15 × 10-10 ). For overlap between imaging-defined and cognitively defined subgroups, AD-Memory matched up with an imaging-defined limbic predominant group. DISCUSSION: MRI findings differ across cognitively defined AD subgroups.

A cross-sectional study of α-synuclein seed amplification assay in Alzheimer's disease neuroimaging initiative: Prevalence and associations with Alzheimer's disease biomarkers and cognitive function.

INTRODUCTION: Alzheimer's disease (AD) pathology is defined by ß-amyloid (Aß) plaques and neurofibrillary tau, but Lewy bodies (LBs; 𝛼-synuclein aggregates) are a common co-pathology for which effective biomarkers are needed. METHODS: A validated α-synuclein Seed Amplification Assay (SAA) was used on recent cerebrospinal fluid (CSF) samples from 1638 Alzheimer's Disease Neuroimaging Initiative (ADNI) participants, 78 with LB-pathology confirmation at autopsy. We compared SAA outcomes with neuropathology, Aß and tau biomarkers, risk-factors, genetics, and cognitive trajectories. RESULTS: SAA showed 79% sensitivity and 97% specificity for LB pathology, with superior performance in identifying neocortical (100%) compared to limbic (57%) and amygdala-predominant (60%) LB-pathology. SAA+ rate was 22%, increasing with disease stage and age. Higher Aß burden but lower CSF p-tau181 associated with higher SAA+ rates, especially in dementia. SAA+ affected cognitive impairment in MCI and Early-AD who were already AD biomarker positive. DISCUSSION: SAA is a sensitive, specific marker for LB-pathology. Its increase in prevalence with age and AD stages, and its association with AD biomarkers, highlights the clinical importance of α-synuclein co-pathology in understanding AD's nature and progression. HIGHLIGHTS: SAA shows 79% sensitivity, 97% specificity for LB-pathology detection in AD. SAA positivity prevalence increases with disease stage and age. Higher Aß burden, lower CSF p-tau181 linked with higher SAA+ rates in dementia. SAA+ impacts cognitive impairment in early disease stages. Study underpins need for wider LB-pathology screening in AD treatment.

Sex-specific genetic architecture of late-life memory performance.

BACKGROUND: Women demonstrate a memory advantage when cognitively healthy yet lose this advantage to men in Alzheimer's disease. However, the genetic underpinnings of this sex difference in memory performance remain unclear. METHODS: We conducted the largest sex-aware genetic study on late-life memory to date (Nmales  = 11,942; Nfemales  = 15,641). Leveraging harmonized memory composite scores from four cohorts of cognitive aging and AD, we performed sex-stratified and sex-interaction genome-wide association studies in 24,216 non-Hispanic White and 3367 non-Hispanic Black participants. RESULTS: We identified three sex-specific loci (rs67099044-CBLN2, rs719070-SCHIP1/IQCJ-SCHIP), including an X-chromosome locus (rs5935633-EGL6/TCEANC/OFD1), that associated with memory. Additionally, we identified heparan sulfate signaling as a sex-specific pathway and found sex-specific genetic correlations between memory and cardiovascular, immune, and education traits. DISCUSSION: This study showed memory is highly and comparably heritable across sexes, as well as highlighted novel sex-specific genes, pathways, and genetic correlations that related to late-life memory. HIGHLIGHTS: Demonstrated the heritable component of late-life memory is similar across sexes. Identified two genetic loci with a sex-interaction with baseline memory. Identified an X-chromosome locus associated with memory decline in females. Highlighted sex-specific candidate genes and pathways associated with memory. Revealed sex-specific shared genetic architecture between memory and complex traits.

Longitudinal change in memory performance as a strong endophenotype for Alzheimer's disease.

INTRODUCTION: Although large-scale genome-wide association studies (GWAS) have been conducted on AD, few have been conducted on continuous measures of memory performance and memory decline. METHODS: We conducted a cross-ancestry GWAS on memory performance (in 27,633 participants) and memory decline (in 22,365 participants; 129,201 observations) by leveraging harmonized cognitive data from four aging cohorts. RESULTS: We found high heritability for two ancestry backgrounds. Further, we found a novel ancestry locus for memory decline on chromosome 4 (rs6848524) and three loci in the non-Hispanic Black ancestry group for memory performance on chromosomes 2 (rs111471504), 7 (rs4142249), and 15 (rs74381744). In our gene-level analysis, we found novel genes for memory decline on chromosomes 1 (SLC25A44), 11 (BSX), and 15 (DPP8). Memory performance and memory decline shared genetic architecture with AD-related traits, neuropsychiatric traits, and autoimmune traits. DISCUSSION: We discovered several novel loci, genes, and genetic correlations associated with late-life memory performance and decline. HIGHLIGHTS: Late-life memory has high heritability that is similar across ancestries. We discovered four novel variants associated with late-life memory. We identified four novel genes associated with late-life memory. Late-life memory shares genetic architecture with psychiatric/autoimmune traits.

Presenilin-1 mutation position influences amyloidosis, small vessel disease, and dementia with disease stage.

INTRODUCTION: Amyloidosis, including cerebral amyloid angiopathy, and markers of small vessel disease (SVD) vary across dominantly inherited Alzheimer's disease (DIAD) presenilin-1 (PSEN1) mutation carriers. We investigated how mutation position relative to codon 200 (pre-/postcodon 200) influences these pathologic features and dementia at different stages. METHODS: Individuals from families with known PSEN1 mutations (n = 393) underwent neuroimaging and clinical assessments. We cross-sectionally evaluated regional Pittsburgh compound B-positron emission tomography uptake, magnetic resonance imaging markers of SVD (diffusion tensor imaging-based white matter injury, white matter hyperintensity volumes, and microhemorrhages), and cognition. RESULTS: Postcodon 200 carriers had lower amyloid burden in all regions but worse markers of SVD and worse Clinical Dementia Rating® scores compared to precodon 200 carriers as a function of estimated years to symptom onset. Markers of SVD partially mediated the mutation position effects on clinical measures. DISCUSSION: We demonstrated the genotypic variability behind spatiotemporal amyloidosis, SVD, and clinical presentation in DIAD, which may inform patient prognosis and clinical trials. HIGHLIGHTS: Mutation position influences Aß burden, SVD, and dementia. PSEN1 pre-200 group had stronger associations between Aß burden and disease stage. PSEN1 post-200 group had stronger associations between SVD markers and disease stage. PSEN1 post-200 group had worse dementia score than pre-200 in late disease stage. Diffusion tensor imaging-based SVD markers mediated mutation position effects on dementia in the late stage.

Brain-wide genome-wide colocalization study for integrating genetics, transcriptomics and brain morphometry in Alzheimer's disease.

Alzheimer's disease (AD) is one of the most common neurodegenerative diseases. However, the AD mechanism has not yet been fully elucidated to date, hindering the development of effective therapies. In our work, we perform a brain imaging genomics study to link genetics, single-cell gene expression data, tissue-specific gene expression data, brain imaging-derived volumetric endophenotypes, and disease diagnosis to discover potential underlying neurobiological pathways for AD. To do so, we perform brain-wide genome-wide colocalization analyses to integrate multidimensional imaging genomic biobank data. Specifically, we use (1) the individual-level imputed genotyping data and magnetic resonance imaging (MRI) data from the UK Biobank, (2) the summary statistics of the genome-wide association study (GWAS) from multiple European ancestry cohorts, and (3) the tissue-specific cis-expression quantitative trait loci (cis-eQTL) summary statistics from the GTEx project. We apply a Bayes factor colocalization framework and mediation analysis to these multi-modal imaging genomic data. As a result, we derive the brain regional level GWAS summary statistics for 145 brain regions with 482,831 single nucleotide polymorphisms (SNPs) followed by posthoc functional annotations. Our analysis yields the discovery of a potential AD causal pathway from a systems biology perspective: the SNP chr10:124165615:G>A (rs6585827) mutation upregulates the expression of BTBD16 gene in oligodendrocytes, a specialized glial cells, in the brain cortex, leading to a reduced risk of volumetric loss in the entorhinal cortex, resulting in the protective effect on AD. We substantiate our findings with multiple evidence from existing imaging, genetic and genomic studies in AD literature. Our study connects genetics, molecular and cellular signatures, regional brain morphologic endophenotypes, and AD diagnosis, providing new insights into the mechanistic understanding of the disease. Our findings can provide valuable guidance for subsequent therapeutic target identification and drug discovery in AD.

Epigenetic aging in older breast cancer survivors and noncancer controls: preliminary findings from the Thinking and Living with Cancer Study.

BACKGROUND: Cancer and its treatments may accelerate aging in survivors; however, research has not examined epigenetic markers of aging in longer term breast cancer survivors. This study examined whether older breast cancer survivors showed greater epigenetic aging than noncancer controls and whether epigenetic aging related to functional outcomes. METHODS: Nonmetastatic breast cancer survivors (n = 89) enrolled prior to systemic therapy and frequency-matched controls (n = 101) ages 62 to 84 years provided two blood samples to derive epigenetic aging measures (Horvath, Extrinsic Epigenetic Age [EEA], PhenoAge, GrimAge, Dunedin Pace of Aging) and completed cognitive (Functional Assessment of Cancer Therapy-Cognitive Function) and physical (Medical Outcomes Study Short Form-12) function assessments at approximately 24 to 36 and 60 months after enrollment. Mixed-effects models tested survivor-control differences in epigenetic aging, adjusting for age and comorbidities; models for functional outcomes also adjusted for racial group, site, and cognitive reserve. RESULTS: Survivors were 1.04 to 2.22 years biologically older than controls on Horvath, EEA, GrimAge, and DunedinPACE measures (p = .001-.04) at approximately 24 to 36 months after enrollment. Survivors exposed to chemotherapy were 1.97 to 2.71 years older (p = .001-.04), and among this group, an older EEA related to worse self-reported cognition (p = .047) relative to controls. An older epigenetic age related to worse physical function in all women (p < .001-.01). Survivors and controls showed similar epigenetic aging over time, but Black survivors showed accelerated aging over time relative to non-Hispanic White survivors. CONCLUSION: Older breast cancer survivors, particularly those exposed to chemotherapy, showed greater epigenetic aging than controls that may relate to worse outcomes. If replicated, measurement of biological aging could complement geriatric assessments to guide cancer care for older women.

Plasma levels of interleukin-6 mediate neurocognitive performance in older breast cancer survivors: The Thinking and Living With Cancer study.

BACKGROUND: Immune activation/inflammation markers (immune markers) were tested to explain differences in neurocognition among older breast cancer survivors versus noncancer controls. METHODS: Women >60 years old with primary breast cancer (stages 0-III) (n = 400) were assessed before systemic therapy with frequency-matched controls (n = 329) and followed annually to 60 months; blood was collected during annual assessments from 2016 to 2020. Neurocognition was measured by tests of attention, processing speed, and executive function (APE). Plasma levels of interleukin-6 (IL-6), IL-8, IL-10, tumor necrosis factor α (TNF-α), and interferon γ were determined using multiplex testing. Mixed linear models were used to compare results of immune marker levels by survivor/control group by time and by controlling for age, racial/ethnic group, cognitive reserve, and study site. Covariate-adjusted multilevel mediation analyses tested whether survivor/control group effects on cognition were explained by immune markers; secondary analyses examined the impact of additional covariates (e.g., comorbidity and obesity) on mediation effects. RESULTS: Participants were aged 60-90 years (mean, 67.7 years). Most survivors had stage I (60.9%) estrogen receptor-positive tumors (87.6%). Survivors had significantly higher IL-6 levels than controls before systemic therapy and at 12, 24, and 60 months (p ≤ .001-.014) but there were no differences for other markers. Survivors had lower adjusted APE scores than controls (p < .05). Levels of IL-6, IL-10, and TNF-α were related to APE, with IL-6 explaining part of the relationship between survivor/control group and APE (p = .01). The magnitude of this mediation effect decreased but remained significant (p = .047) after the consideration of additional covariates. CONCLUSIONS: Older breast cancer survivors had worse long-term neurocognitive performance than controls, and this relationship was explained in part by elevated IL-6.

A novel deep learning method for predictive modeling of microbiome data.

With the development and decreasing cost of next-generation sequencing technologies, the study of the human microbiome has become a rapid expanding research field, which provides an unprecedented opportunity in various clinical applications such as drug response predictions and disease diagnosis. It is thus essential and desirable to build a prediction model for clinical outcomes based on microbiome data that usually consist of taxon abundance and a phylogenetic tree. Importantly, all microbial species are not uniformly distributed in the phylogenetic tree but tend to be clustered at different phylogenetic depths. Therefore, the phylogenetic tree represents a unique correlation structure of microbiome, which can be an important prior to improve the prediction performance. However, prediction methods that consider the phylogenetic tree in an efficient and rigorous way are under-developed. Here, we develop a novel deep learning prediction method MDeep (microbiome-based deep learning method) to predict both continuous and binary outcomes. Conceptually, MDeep designs convolutional layers to mimic taxonomic ranks with multiple convolutional filters on each convolutional layer to capture the phylogenetic correlation among microbial species in a local receptive field and maintain the correlation structure across different convolutional layers via feature mapping. Taken together, the convolutional layers with its built-in convolutional filters capture microbial signals at different taxonomic levels while encouraging local smoothing and preserving local connectivity induced by the phylogenetic tree. We use both simulation studies and real data applications to demonstrate that MDeep outperforms competing methods in both regression and binary classifications. Availability and Implementation: MDeep software is available at https://github.com/lichen-lab/MDeep Contact:chen61@iu.edu.

Integrative-omics for discovery of network-level disease biomarkers: a case study in Alzheimer's disease.

A large number of genetic variations have been identified to be associated with Alzheimer's disease (AD) and related quantitative traits. However, majority of existing studies focused on single types of omics data, lacking the power of generating a community including multi-omic markers and their functional connections. Because of this, the immense value of multi-omics data on AD has attracted much attention. Leveraging genomic, transcriptomic and proteomic data, and their backbone network through functional relations, we proposed a modularity-constrained logistic regression model to mine the association between disease status and a group of functionally connected multi-omic features, i.e. single-nucleotide polymorphisms (SNPs), genes and proteins. This new model was applied to the real data collected from the frontal cortex tissue in the Religious Orders Study and Memory and Aging Project cohort. Compared with other state-of-art methods, it provided overall the best prediction performance during cross-validation. This new method helped identify a group of densely connected SNPs, genes and proteins predictive of AD status. These SNPs are mostly expression quantitative trait loci in the frontal region. Brain-wide gene expression profile of these genes and proteins were highly correlated with the brain activation map of 'vision', a brain function partly controlled by frontal cortex. These genes and proteins were also found to be associated with the amyloid deposition, cortical volume and average thickness of frontal regions. Taken together, these results suggested a potential pathway underlying the development of AD from SNPs to gene expression, protein expression and ultimately brain functional and structural changes.

WEVar: a novel statistical learning framework for predicting noncoding regulatory variants.

Understanding the functional consequence of noncoding variants is of great interest. Though genome-wide association studies or quantitative trait locus analyses have identified variants associated with traits or molecular phenotypes, most of them are located in the noncoding regions, making the identification of causal variants a particular challenge. Existing computational approaches developed for prioritizing noncoding variants produce inconsistent and even conflicting results. To address these challenges, we propose a novel statistical learning framework, which directly integrates the precomputed functional scores from representative scoring methods. It will maximize the usage of integrated methods by automatically learning the relative contribution of each method and produce an ensemble score as the final prediction. The framework consists of two modes. The first 'context-free' mode is trained using curated causal regulatory variants from a wide range of context and is applicable to predict regulatory variants of unknown and diverse context. The second 'context-dependent' mode further improves the prediction when the training and testing variants are from the same context. By evaluating the framework via both simulation and empirical studies, we demonstrate that it outperforms integrated scoring methods and the ensemble score successfully prioritizes experimentally validated regulatory variants in multiple risk loci.