Temporal tau asymmetry spectrum influences divergent behavior and language patterns in Alzheimer's disease.

Understanding the psychiatric symptoms of Alzheimer s disease (AD) is crucial for advancing precision medicine and therapeutic strategies. The relationship between AD behavioral symptoms and asymmetry in spatial tau PET patterns is not well-known. Braak tau progression implicates the temporal lobes early. However, the clinical and pathological implications of temporal tau laterality remain unexplored. This cross-sectional study investigated the correlation between temporal tau PET asymmetry and behavior assessed using the neuropsychiatric inventory and composite scores for memory, executive function, and language, using data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. In the entire cohort, continuous right and left temporal tau contributions to behavior and cognition were evaluated, controlling for age, sex, education, and tau burden on the contralateral side. Additionally, a temporal tau laterality index was calculated to define "asymmetry-extreme" groups (individuals with laterality indices greater than two standard deviations from the mean). 695 individuals (age = 73.9 ± 7.6 years, 372 (53.5 %) females) were included, comprising 281 (40%) cognitively unimpaired (CU) amyloid negative, 185 (27%) CU amyloid positive, and 229 (33%) impaired (CI) amyloid positive participants. In the full cohort analysis, right temporal tau was associated with worse behavior (B = 8.14, p-value = 0.007), and left temporal tau was associated with worse language (B = 1.4, p-value < 0.001). Categorization into asymmetry-extreme groups revealed 20 right- and 27 left-asymmetric participants. Within these extreme groups, there was additional heterogeneity along the anterior-posterior dimension. Asymmetrical tau burden is associated with distinct behavioral and cognitive profiles. Wide multi-cultural implementation of social cognition measures is needed to understand right-sided asymmetry in AD.

Stability of transcranial magnetic stimulation electroencephalogram evoked potentials in pediatric epilepsy.

Transcranial magnetic stimulation paired with electroencephalography (TMS-EEG) can measure local excitability and functional connectivity. To address trial-to-trial variability, responses to multiple TMS pulses are recorded to obtain an average TMS evoked potential (TEP). Balancing adequate data acquisition to establish stable TEPs with feasible experimental duration is critical when applying TMS-EEG to clinical populations. Here we aim to investigate the minimum number of pulses (MNP) required to achieve stable TEPs in children with epilepsy. Eighteen children with Self-Limited Epilepsy with Centrotemporal Spikes, a common epilepsy arising from the motor cortices, underwent multiple 100-pulse blocks of TMS to both motor cortices over two days. TMS was applied at 120% of resting motor threshold (rMT) up to a maximum of 100% maximum stimulator output. The average of all 100 pulses was used as a "gold-standard" TEP to which we compared "candidate" TEPs obtained by averaging subsets of pulses. We defined TEP stability as the MNP needed to achieve a concordance correlation coefficient of 80% between the candidate and "gold-standard" TEP. We additionally assessed whether experimental or clinical factors affected TEP stability. Results show that stable TEPs can be derived from fewer than 100 pulses, a number typically used for designing TMS-EEG experiments. The early segment (15-80 ms) of the TEP was less stable than the later segment (80-350 ms). Global mean field amplitude derived from all channels was less stable than local TEP derived from channels overlying the stimulated site. TEP stability did not differ depending on stimulated hemisphere, block order, or antiseizure medication use, but was greater in older children. Stimulation administered with an intensity above the rMT yielded more stable local TEPs. Studies of TMS-EEG in pediatrics have been limited by the complexity of experimental set-up and time course. This study serves as a critical starting point, demonstrating the feasibility of designing efficient TMS-EEG studies that use a relatively small number of pulses to study pediatric epilepsy and potentially other pediatric groups.

Beyond guilty by association at scale: searching for causal variants on the basis of genome-wide summary statistics.

Understanding the causal genetic architecture of complex phenotypes is essential for future research into disease mechanisms and potential therapies. Here, we present a novel framework for genome-wide detection of sets of variants that carry non-redundant information on the phenotypes and are therefore more likely to be causal in a biological sense. Crucially, our framework requires only summary statistics obtained from standard genome-wide marginal association testing. The described approach, implemented in open-source software, is also computationally efficient, requiring less than 15 minutes on a single CPU to perform genome-wide analysis. Through extensive genome-wide simulation studies, we show that the method can substantially outperform usual two-stage marginal association testing and fine-mapping procedures in precision and recall. In applications to a meta-analysis of ten large-scale genetic studies of Alzheimer's disease (AD), we identified 82 loci associated with AD, including 37 additional loci missed by conventional GWAS pipeline. The identified putative causal variants achieve state-of-the-art agreement with massively parallel reporter assays and CRISPR-Cas9 experiments. Additionally, we applied the method to a retrospective analysis of 67 large-scale GWAS summary statistics since 2013 for a variety of phenotypes. Results reveal the method's capacity to robustly discover additional loci for polygenic traits and pinpoint potential causal variants underpinning each locus beyond conventional GWAS pipeline, contributing to a deeper understanding of complex genetic architectures in post-GWAS analyses.

Second-order group knockoffs with applications to GWAS.

Conditional testing via the knockoff framework allows one to identify -- among large number of possible explanatory variables -- those that carry unique information about an outcome of interest, and also provides a false discovery rate guarantee on the selection. This approach is particularly well suited to the analysis of genome wide association studies (GWAS), which have the goal of identifying genetic variants which influence traits of medical relevance. While conditional testing can be both more powerful and precise than traditional GWAS analysis methods, its vanilla implementation encounters a difficulty common to all multivariate analysis methods: it is challenging to distinguish among multiple, highly correlated regressors. This impasse can be overcome by shifting the object of inference from single variables to groups of correlated variables. To achieve this, it is necessary to construct "group knockoffs." While successful examples are already documented in the literature, this paper substantially expands the set of algorithms and software for group knockoffs. We focus in particular on second-order knockoffs, for which we describe correlation matrix approximations that are appropriate for GWAS data and that result in considerable computational savings. We illustrate the effectiveness of the proposed methods with simulations and with the analysis of albuminuria data from the UK Biobank. The described algorithms are implemented in an open-source Julia package Knockoffs.jl, for which both R and Python wrappers are available.

Controlled Variable Selection from Summary Statistics Only? A Solution via GhostKnockoffs and Penalized Regression.

Identifying which variables do influence a response while controlling false positives pervades statistics and data science. In this paper, we consider a scenario in which we only have access to summary statistics, such as the values of marginal empirical correlations between each dependent variable of potential interest and the response. This situation may arise due to privacy concerns, e.g., to avoid the release of sensitive genetic information. We extend GhostKnockoffs He et al. [2022] and introduce variable selection methods based on penalized regression achieving false discovery rate (FDR) control. We report empirical results in extensive simulation studies, demonstrating enhanced performance over previous work. We also apply our methods to genome-wide association studies of Alzheimer's disease, and evidence a significant improvement in power.

Long-term persistence to OnabotulinumtoxinA to prevent chronic migraine: Results from 11 years of patient data from a tertiary headache center.

OBJECTIVE: To determine if patients with chronic migraine continue onabotulinumtoxinA (onabotA) long-term. METHODS: We performed a retrospective cohort analysis using aggregated, de-identified patient data from the Stanford Headache Center. We included patients in California who received at least one prescription for onabotA during the years of 2011-2021. The primary outcome was the number of onabotA treatments each patient received. Secondary outcomes included sex, age, race, ethnicity, body mass index (BMI), distance to the treatment facility, and zip code income quartile. RESULTS: A total of 1,551 patients received a mean of 7.60 ± 7.26 treatments and a median of 5 treatments, with 16.2% of patients receiving only one treatment and 10.6% receiving at least 19. Time-to-event survival analysis suggested 26.0% of patients would complete at least 29 treatments if able. Younger age and female sex were associated with statistically significant differences between quartile groups of number of onabotA treatments (p = 0.007, p = 0.015). BMI, distance to treatment facility, and zip code income quartile were not statistically significantly different between quartile groups (p > 0.500 for all). Prescriptions of both triptans and non-onabotA preventive medications showed a statistically significant increase with each higher quartile of number of onabotA treatments (p < 0.001; p < 0.001). DISCUSSION: We show long-term persistence to onabotA is high and that distance to treatment facility and income are not factors in continuation. Our work also demonstrates that as patients continue onabotA over time, there may be an increased need for adjunctive or alternative treatments.

Impaired 24-h activity patterns are associated with an increased risk of Alzheimer's disease, Parkinson's disease, and cognitive decline.

BACKGROUND: Sleep-wake regulating circuits are affected during prodromal stages in the pathological progression of both Alzheimer's disease (AD) and Parkinson's disease (PD), and this disturbance can be measured passively using wearable devices. Our objective was to determine whether accelerometer-based measures of 24-h activity are associated with subsequent development of AD, PD, and cognitive decline. METHODS: This study obtained UK Biobank data from 82,829 individuals with wrist-worn accelerometer data aged 40 to 79 years with a mean (± SD) follow-up of 6.8 (± 0.9) years. Outcomes were accelerometer-derived measures of 24-h activity (derived by cosinor, nonparametric, and functional principal component methods), incident AD and PD diagnosis (obtained through hospitalization or primary care records), and prospective longitudinal cognitive testing. RESULTS: One hundred eighty-seven individuals progressed to AD and 265 to PD. Interdaily stability (a measure of regularity, hazard ratio [HR] per SD increase 1.25, 95% confidence interval [CI] 1.05-1.48), diurnal amplitude (HR 0.79, CI 0.65-0.96), mesor (mean activity; HR 0.77, CI 0.59-0.998), and activity during most active 10 h (HR 0.75, CI 0.61-0.94), were associated with risk of AD. Diurnal amplitude (HR 0.28, CI 0.23-0.34), mesor (HR 0.13, CI 0.10-0.16), activity during least active 5 h (HR 0.24, CI 0.08-0.69), and activity during most active 10 h (HR 0.20, CI 0.16-0.25) were associated with risk of PD. Several measures were additionally predictive of longitudinal cognitive test performance. CONCLUSIONS: In this community-based longitudinal study, accelerometer-derived metrics were associated with elevated risk of AD, PD, and accelerated cognitive decline. These findings suggest 24-h rhythm integrity, as measured by affordable, non-invasive wearable devices, may serve as a scalable early marker of neurodegenerative disease.

The introduction of the CGRP monoclonal antibodies and their effect on the prescription patterns of chronic migraine preventive medications in a tertiary headache center: A retrospective, observational analysis.

OBJECTIVE: To determine the effect of the introduction of the calcitonin gene-related peptide monoclonal antibodies (CGRP mAbs) in 2018 on the prescribing of older medications for the prevention of chronic migraine. BACKGROUND: Prior to 2018, the preventive treatment of migraine borrowed from medications intended to treat other illnesses with the last medication, onabotulinumtoxinA, receiving Food and Drug Administration (FDA) approval for the prevention of chronic migraine in 2010. The FDA approval of three CGRP mAbs in 2018 provided the ideal natural experiment to assess how the introduction of these medications, and a fourth in 2020, affected the generally stable migraine preventive medications market. METHODS: We performed a retrospective cohort analysis using the aggregated de-identified data of 6595 patients. The percentage of patients with chronic migraine who had been prescribed one of ten most prescribed oral preventive medications or onabotulinumtoxinA, or any of the four CGRP mAbs, were calculated relative to the total number of patients with chronic migraine who received a prescription for any medication from our clinic during the pre-CGRP mAb years of 2015-2017 and post-approval years of 2019-2021. RESULTS: We observed a statistically significant decrease in the prescription of the top 10 most prescribed medications after the introduction of the CGRP mAbs overall (1456/3144, 46.3%, to 1995/4629, 43.1%, p = 0.001), as well as with most individual medications, including large decreases in verapamil (230/3144, 7.3%, to 125/4629, 2.7%; p < 0.001), the tricyclic antidepressants (494/3144, 15.7%, to 532/4629, 11.5%; p < 0.001), topiramate (566/3144, 18.0%, to 653/4629, 14.1%; p < 0.001), and onabotulinumtoxinA (861/3144, 27.4%, to 1134/4629, 24.5%; p = 0.001). CONCLUSION: The introduction of the CGRP mAbs during 2018 resulted in a decrease in utilization of most oral medications and onabotulinumtoxinA for the prevention of migraine. Future work should continue to observe how the prescription patterns of these medications evolve with time.

Post-translational modifications linked to preclinical Alzheimer's disease-related pathological and cognitive changes.

INTRODUCTION: In this study, we leverage proteomic techniques to identify communities of proteins underlying Alzheimer's disease (AD) risk among clinically unimpaired (CU) older adults. METHODS: We constructed a protein co-expression network using 3869 cerebrospinal fluid (CSF) proteins quantified by SomaLogic, Inc., in a cohort of participants along the AD clinical spectrum. We then replicated this network in an independent cohort of CU older adults and related these modules to clinically-relevant outcomes. RESULTS: We discovered modules enriched for phosphorylation and ubiquitination that were associated with abnormal amyloid status, as well as p-tau181 (M4: ß = 2.44, p < 0.001, M7: ß = 2.57, p < 0.001) and executive function performance (M4: ß = -2.00, p = 0.005, M7: ß = -2.39, p < 0.001). DISCUSSION: In leveraging CSF proteomic data from individuals spanning the clinical spectrum of AD, we highlight the importance of post-translational modifications for early cognitive and pathological changes.

Organ aging signatures in the plasma proteome track health and disease.

Animal studies show aging varies between individuals as well as between organs within an individual1-4, but whether this is true in humans and its effect on age-related diseases is unknown. We utilized levels of human blood plasma proteins originating from specific organs to measure organ-specific aging differences in living individuals. Using machine learning models, we analysed aging in 11 major organs and estimated organ age reproducibly in five independent cohorts encompassing 5,676 adults across the human lifespan. We discovered nearly 20% of the population show strongly accelerated age in one organ and 1.7% are multi-organ agers. Accelerated organ aging confers 20-50% higher mortality risk, and organ-specific diseases relate to faster aging of those organs. We find individuals with accelerated heart aging have a 250% increased heart failure risk and accelerated brain and vascular aging predict Alzheimer's disease (AD) progression independently from and as strongly as plasma pTau-181 (ref. 5), the current best blood-based biomarker for AD. Our models link vascular calcification, extracellular matrix alterations and synaptic protein shedding to early cognitive decline. We introduce a simple and interpretable method to study organ aging using plasma proteomics data, predicting diseases and aging effects.

Assessing the Assisted Six-Minute Cycling Test as a Measure of Endurance in Non-Ambulatory Patients with Spinal Muscular Atrophy (SMA).

Assessing endurance in non-ambulatory individuals with Spinal Muscular Atrophy (SMA) has been challenging due to limited evaluation tools. The Assisted 6-Minute Cycling Test (A6MCT) is an upper limb ergometer assessment used in other neurologic disorders to measure endurance. To study the performance of the A6MCT in the non-ambulatory SMA population, prospective data was collected on 38 individuals with SMA (13 sitters; 25 non-sitters), aged 5 to 74 years (mean = 30.3; SD = 14.1). The clinical measures used were A6MCT, Revised Upper Limb Module (RULM), Adapted Test of Neuromuscular Disorders (ATEND), and Egen Klassifikation Scale 2 (EK2). Perceived fatigue was assessed using the Fatigue Severity Scale (FSS), and effort was assessed using the Rate of Perceived Exertion (RPE). Data were analyzed for: (1) Feasibility, (2) Clinical discrimination, and (3) Associations between A6MCT with clinical characteristics and outcomes. Results showed the A6MCT was feasible for 95% of the tested subjects, discriminated between functional groups (p = 0.0086), and was significantly associated with results obtained from RULM, ATEND, EK2, and Brooke (p < 0.0001; p = 0.029; p < 0.001; p = 0.005). These findings indicate the A6MCT's potential to evaluate muscular endurance in non-ambulatory SMA individuals, complementing clinician-rated assessments. Nevertheless, further validation with a larger dataset is needed for broader application.

Temporal tau asymmetry spectrum influences divergent behavior and language patterns in Alzheimer`s disease.

Understanding psychiatric symptoms in Alzheimer`s disease (AD) is crucial for advancing precision medicine and therapeutic strategies. The relationship between AD behavioral symptoms and asymmetry in spatial tau PET patterns is unknown. Braak tau progression implicates the temporal lobes early. However, the clinical and pathological implications of temporal tau laterality remain unexplored. This cross-sectional study investigated the correlation between temporal tau PET asymmetry and behavior assessed using the neuropsychiatric inventory, and composite scores for memory, executive function, and language; using data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. In the entire cohort, continuous right and left temporal tau contributions to behavior and cognition were evaluated controlling for age, sex, education, and tau burden on the contralateral side. Additionally, a temporal tau laterality index was calculated to define "asymmetry-extreme" groups (individuals with laterality indices greater than two standard deviations from the mean). 858 individuals (age=73.9±7.7 years, 434(50%) females) were included, comprising 438 cognitively unimpaired (CU) (53.4%) and 420 impaired (CI) participants (48.9%). In the full cohort analysis, right temporal tau was associated with worse behavior (B(SE)=7.19 (2.9), p-value=0.01) and left temporal tau was associated with worse language (B(SE)=1.4(0.2), p-value<0.0001). Categorization into asymmetry-extreme groups revealed 20 right- and 27 left-asymmetric participants. Within these extreme groups, four patterns of tau PET uptake were observed: anterior temporal, typical AD, typical AD with frontal involvement, and posterior. Asymmetrical tau burden is associated with distinct behavioral and cognitive profiles. Behavioral and socioemotional measures are needed to understand right-sided asymmetry in AD.

Major Adverse Dystrophinopathy Events (MADE) Score as Marker of Cumulative Morbidity and Risk for Mortality in Boys with Duchenne Muscular Dystrophy.

Background: Overlapping symptoms from cardiomyopathy, respiratory insufficiency, and skeletal myopathy confound assessment of heart failure in Duchenne Muscular Dystrophy. We developed an ordinal scale of multiorgan clinical variables that reflect cumulative disease burden-the Major Adverse Dystrophinopathy Event (MADE) Score. We hypothesized that a higher MADE score would be associated with increased mortality in boys with Duchenne Muscular Dystrophy. The Cooperative International Neuromuscular Research Group Duchenne Natural History Study dataset was utilized for validation. Methods: Duchenne Natural History Study variables were selected based on clinical relevance to prespecified domains: Cardiac, Pulmonary, Myopathy, Nutrition. Severity points (0-4) were assigned and summed for study visits. MADE score for cohorts defined by age, ambulatory status, and survival were compared at enrollment and longitudinally.Associations between MADE score and mortality were examined. Results: Duchenne Natural History Study enrolled 440 males, 12.6 ±6.1 years old, with 3,559 visits over 4.6 ±2.8 years, 45 deaths. MADE score increased with age and nonambulatory status. Mean MADE score per visit was 19 ±10 for those who died vs. 9.8 ±9.3 in survivors p=0.03. Baseline MADE score >12 predicted mortality independent of age (78% sensitivity, CPE.70). Rising MADE score trajectory was associated with mortality in models adjusted for enrollment age, follow-up time, and ambulatory status, all p<.001. Conclusion: A multiorgan severity score, MADE, was developed to track cumulative morbidities that impact heart failure in Duchenne muscular dystrophy. MADE score predicted Duchenne Natural History Study mortality. MADE score can be used for serial heart failure assessment in males and may serve as an endpoint for Duchenne muscular dystrophy clinical research.

Loop diuretics association with Alzheimer's disease risk.

Objectives: To investigate whether exposure history to two common loop diuretics, bumetanide and furosemide, affects the risk of developing Alzheimer's disease (AD) after accounting for socioeconomic status and congestive heart failure. Methods: Individuals exposed to bumetanide or furosemide were identified in the Stanford University electronic health record using the de-identified Observational Medical Outcomes Partnership platform. We matched the AD case cohort to a control cohort (1:20 case:control) on gender, race, ethnicity, and hypertension, and controlled for variables that could potentially be collinear with bumetanide exposure and/or AD diagnosis. Among individuals older than 65 years, 5,839 AD cases and 116,103 matched controls were included. A total of 1,759 patients (54 cases and 1,705 controls) were exposed to bumetanide. Results: After adjusting for socioeconomic status and other confounders, the exposure of bumetanide and furosemide was significantly associated with reduced AD risk (respectively, bumetanide odds ratio [OR] = 0.23; 95% confidence interval [CI], 0.15-0.36; p = 4.0 × 10-11; furosemide OR = 0.42; 95% CI, 0.38-0.47; p < 2.0 × 10-16). Discussion: Our study replicates in an independent sample that a history of bumetanide exposure is associated with reduced AD risk while also highlighting an association of the most common loop diuretic (furosemide) with reduced AD risk. These associations need to be additionally replicated, and the mechanism of action remains to be investigated.

Improving genetic risk prediction across diverse population by disentangling ancestry representations.

Risk prediction models using genetic data have seen increasing traction in genomics. However, most of the polygenic risk models were developed using data from participants with similar (mostly European) ancestry. This can lead to biases in the risk predictors resulting in poor generalization when applied to minority populations and admixed individuals such as African Americans. To address this issue, largely due to the prediction models being biased by the underlying population structure, we propose a deep-learning framework that leverages data from diverse population and disentangles ancestry from the phenotype-relevant information in its representation. The ancestry disentangled representation can be used to build risk predictors that perform better across minority populations. We applied the proposed method to the analysis of Alzheimer's disease genetics. Comparing with standard linear and nonlinear risk prediction methods, the proposed method substantially improves risk prediction in minority populations, including admixed individuals, without needing self-reported ancestry information.

Bumetanide Exposure Association with Alzheimer's Disease Risk.

Background: To investigate whether exposure history to two common loop diuretics affects the risk of developing Alzheimer's disease (AD) after accounting for socioeconomic status and congestive heart failure. Methods: Individuals exposed to bumetanide or furosemide were identified in the Stanford University electronic health record using the deidentified Observational Medical Outcomes Partnership platform. We matched the AD case cohort to a control cohort (1:20 case:control) on gender, race, ethnicity, hypertension and controlled for variables that could potentially be collinear with bumetanide exposure and/or AD diagnosis. Among individuals older than 65 years, 5,839 AD cases and 116,103 matched controls were included. A total of 1,759 patients (54 cases, 1,705 controls) were exposed to bumetanide. Results: After adjusting for socioeconomic status and other confounders, bumetanide exposure was significantly associated with reduced AD risk (odds ratio = 0.50; 95% confidence interval, 0.37-0.68; p = 9.9×10-6), while the most common loop diuretics, furosemide, was not associated with AD risk. Conclusion: Our study replicates in an independent sample that history of bumetanide exposure is associated with reduced risk of AD and emphasizes that this association is not confounded by difference in socioeconomic status, which was an important caveat given the cost difference between bumetanide and furosemide.

Association of African Ancestry-Specific APOE Missense Variant R145C With Risk of Alzheimer Disease.

Importance: Numerous studies have established the association of the common APOE ε2 and APOE ε4 alleles with Alzheimer disease (AD) risk across ancestries. Studies of the interaction of these alleles with other amino acid changes on APOE in non-European ancestries are lacking and may improve ancestry-specific risk prediction. Objective: To determine whether APOE amino acid changes specific to individuals of African ancestry modulate AD risk. Design, Setting, and Participants: Case-control study including 31 929 participants and using a sequenced discovery sample (Alzheimer Disease Sequencing Project; stage 1) followed by 2 microarray imputed data sets derived from the Alzheimer Disease Genetic Consortium (stage 2, internal replication) and the Million Veteran Program (stage 3, external validation). This study combined case-control, family-based, population-based, and longitudinal AD cohorts, which recruited participants (1991-2022) in primarily US-based studies with 1 US/Nigerian study. Across all stages, individuals included in this study were of African ancestry. Exposures: Two APOE missense variants (R145C and R150H) were assessed, stratified by APOE genotype. Main Outcomes and Measures: The primary outcome was AD case-control status, and secondary outcomes included age at AD onset. Results: Stage 1 included 2888 cases (median age, 77 [IQR, 71-83] years; 31.3% male) and 4957 controls (median age, 77 [IQR, 71-83] years; 28.0% male). In stage 2, across multiple cohorts, 1201 cases (median age, 75 [IQR, 69-81] years; 30.8% male) and 2744 controls (median age, 80 [IQR, 75-84] years; 31.4% male) were included. In stage 3, 733 cases (median age, 79.4 [IQR, 73.8-86.5] years; 97.0% male) and 19 406 controls (median age, 71.9 [IQR, 68.4-75.8] years; 94.5% male) were included. In ε3/ε4-stratified analyses of stage 1, R145C was present in 52 individuals with AD (4.8%) and 19 controls (1.5%); R145C was associated with an increased risk of AD (odds ratio [OR], 3.01; 95% CI, 1.87-4.85; P = 6.0 × 10-6) and was associated with a reported younger age at AD onset (ß, -5.87 years; 95% CI, -8.35 to -3.4 years; P = 3.4 × 10-6). Association with increased AD risk was replicated in stage 2 (R145C was present in 23 individuals with AD [4.7%] and 21 controls [2.7%]; OR, 2.20; 95% CI, 1.04-4.65; P = .04) and was concordant in stage 3 (R145C was present in 11 individuals with AD [3.8%] and 149 controls [2.7%]; OR, 1.90; 95% CI, 0.99-3.64; P = .051). Association with earlier AD onset was replicated in stage 2 (ß, -5.23 years; 95% CI, -9.58 to -0.87 years; P = .02) and stage 3 (ß, -10.15 years; 95% CI, -15.66 to -4.64 years; P = 4.0 × 10-4). No significant associations were observed in other APOE strata for R145C or in any APOE strata for R150H. Conclusions and Relevance: In this exploratory analysis, the APOE ε3[R145C] missense variant was associated with an increased risk of AD among individuals of African ancestry with the ε3/ε4 genotype. With additional external validation, these findings may inform AD genetic risk assessment in individuals of African ancestry.

BIGKnock: fine-mapping gene-based associations via knockoff analysis of biobank-scale data.

We propose BIGKnock (BIobank-scale Gene-based association test via Knockoffs), a computationally efficient gene-based testing approach for biobank-scale data, that leverages long-range chromatin interaction data, and performs conditional genome-wide testing via knockoffs. BIGKnock can prioritize causal genes over proxy associations at a locus. We apply BIGKnock to the UK Biobank data with 405,296 participants for multiple binary and quantitative traits, and show that relative to conventional gene-based tests, BIGKnock produces smaller sets of significant genes that contain the causal gene(s) with high probability. We further illustrate its ability to pinpoint potential causal genes at [Formula: see text] of the associated loci.

Deep learning-assisted genome-wide characterization of massively parallel reporter assays.

Massively parallel reporter assay (MPRA) is a high-throughput method that enables the study of the regulatory activities of tens of thousands of DNA oligonucleotides in a single experiment. While MPRA experiments have grown in popularity, their small sample sizes compared to the scale of the human genome limits our understanding of the regulatory effects they detect. To address this, we develop a deep learning model, MpraNet, to distinguish potential MPRA targets from the background genome. This model achieves high discriminative performance (AUROC = 0.85) at differentiating MPRA positives from a set of control variants that mimic the background genome when applied to the lymphoblastoid cell line. We observe that existing functional scores represent very distinct functional effects, and most of them fail to characterize the regulatory effect that MPRA detects. Using MpraNet, we predict potential MPRA functional variants across the genome and identify the distributions of MPRA effect relative to other characteristics of genetic variation, including allele frequency, alternative functional annotations specified by FAVOR, and phenome-wide associations. We also observed that the predicted MPRA positives are not uniformly distributed across the genome; instead, they are clumped together in active regions comprising 9.95% of the genome and inactive regions comprising 89.07% of the genome. Furthermore, we propose our model as a screen to filter MPRA experiment candidates at genome-wide scale, enabling future experiments to be more cost-efficient by increasing precision relative to that observed from previous MPRAs.