Research Projects Directory

Deep Learning Derived Phenotypes for Neuro-PASC V7_SHI Billing

We will develop models to predict the risk for neurological complications. In particular, we will use the numeric output that before the softmax in the final binary output as the risk of the predicted neuro-PASC or ADRD-PASC. In addition, we…

Scientific Questions Being Studied

We will develop models to predict the risk for neurological complications. In particular, we will use the numeric output that before the softmax in the final binary output as the risk of the predicted neuro-PASC or ADRD-PASC. In addition, we will also consider the predicted hazards, the output from the survival outcome predictions. Also, to maximize the information captured by our deep learning models, we will also use the neurons in the hidden layer as endophenotypes for neuro-PASC and ADRD-PASC. For all phenotypes, we will conduct single-phenotype GWAS association tests using linear mixed models.

Alzheimer’s disease and related dementia are already an important national priority and Neurological complications of PASC (neuro-PASC) are important as they are a major category of PASC. Finding the related genetic factors behind these two disease can be crucial in the management of early detection.

Project Purpose(s)

Disease Focused Research (COVID-19 and Alzheimer's disease)

Scientific Approaches

Datasets: N3C is one of the richest data sources that include the electronic health records data for more than 5 million confirmed covid-19 patients from 74 sites
across the United States. All of Us is a unique source where we can access the genetic and clinical data for 100000 US patients and with higher representation for
minority groups.

Research Methods and Methods : We train a deep learning-based model on COVID-19 patients’ data available through the N3C initiative. As an outcome, the model will learn a phenotypic representation that consists of the patient's risk to develop post COVID complications, including neuropsychiatric complications Afterwards, we will transfer the model to the All of Us researcher platform and apply our model to create the phenotypic representation for the 11,767 COVID-19 patients using their EHR data. Then, we will use the genotypic data for 3,653 Covid-19 patients who has both their whole genome sequencing (WGS) and EHR data for the GWAS study

Anticipated Findings

The goal is to bring breakthroughs in AI/ML for expedite discovery of the genetic basis of Alzheimer’s disease (AD). We expect to find associations between endophenotypes and SNPs related to Long Covid and Alzheimer's Disease.

The innovation of our project are as follows:
1. Using a transfer learning approach to leverage the large N3C data for phenotyping All of Us data is new.
2. We will be the first who leverage the All of Us platform to study the genetic factors for neuro-PASC and ADRD-PASC.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Controlled Tier

Research Team

Owner:

Laila Rasmy Bekhet - Graduate Trainee, University of Texas Health Science Center, Houston
Hao Yan - Graduate Trainee, University of Texas Health Science Center, Houston
Degui Zhi - Mid-career Tenured Researcher, University of Texas Health Science Center, Houston
Ardalan Naseri - Other, University of Texas Health Science Center, Houston

Alzheimer's disease and related dementias in AOU

We're studying Alzheimer's disease and related dementias using the All of Us cohort and aim to examine association of various variables on dementia such as socio-demographic, genomic, and lifestyle factors through survey questionnaires. This research holds significant importance for both…

Scientific Questions Being Studied

We're studying Alzheimer's disease and related dementias using the All of Us cohort and aim to examine association of various variables on dementia such as socio-demographic, genomic, and lifestyle factors through survey questionnaires. This research holds significant importance for both science and public health. The research can contribute to the development of targeted prevention strategies, personalized treatment approaches, and a deeper understanding of the factors that influence the incidence and progression of these debilitating neurodegenerative conditions.

Project Purpose(s)

Disease Focused Research (Alzheimer's disease)
Population Health
Social / Behavioral
Methods Development
Ancestry

Scientific Approaches

Descriptive Statistics: Conduct initial exploratory analyses to summarize the distribution of socio-demographic variables, genetic variants, and survey responses in the study population. This will help in understanding the characteristics of the cohort and identifying any potential biases.
Correlation and Regression Analysis
Survival Analysis: Employ survival analysis techniques to study the time-to-event data, specifically the time from cohort entry to the development of Alzheimer's disease or dementia-related outcomes. This can provide valuable insights into disease progression and prognosis.
Data Visualization: Create visual representations of the findings to communicate results effectively, such as bar plots, scatter plots, heatmaps, and survival curves.

Anticipated Findings

the anticipated findings from this study have the potential to make significant contributions to the scientific understanding of Alzheimer's disease and related dementias. By shedding light on risk factors, interactions between variables, and predictive models, the research can have practical implications for improving patient outcomes, guiding public health policies, and advancing dementia research and care.

Demographic Categories of Interest

Data Set Used

Controlled Tier

Research Team

Owner:

Minhyuk Choi - Project Personnel, University of California, San Francisco

Collaborators:

Jingxuan Wang - Graduate Trainee, University of California, San Francisco

Dementia and lifestyle correlations

Dementia is a condition characterized by impairment to brain function, often due to diseases such as Alzheimer’s disease or Lewy bodies dementia, and is associated with memory loss and changes in one’s ability to think and function daily. We will…

Scientific Questions Being Studied

Dementia is a condition characterized by impairment to brain function, often due to diseases such as Alzheimer’s disease or Lewy bodies dementia, and is associated with memory loss and changes in one’s ability to think and function daily. We will address the research question of how lifestyle choices affect the development and risk of dementia, with a particular focus on factors such as sleep quality or movement. We aim to affirm and build off possible correlations between certain lifestyle choices and dementia incidence, improve early diagnosis strategies by contributing to existing models, and promote healthy living habits.

Project Purpose(s)

Educational

Scientific Approaches

We plan to carry out data analysis for correlations from the All of Us datasets. We will define our cohorts and control groups of those diagnosed with dementia and those not currently diagnosed with dementia, select possible risk factors, and generate mappings, graphs, and other adjacent functions to analyze correlations under the guidance of our project advisor. We will account for statistical significance when determining the conclusions of our analysis.

Anticipated Findings

We hypothesize that certain lifestyle choices or conditions, e.g., lower-quality sleep, are positively correlated with dementia diagnosis and development. We hope to verify and reference prior studies about dementia risk factors in our analysis, discover new correlations between relatively sparser-covered conditions and their possible correlations with dementia, and analyze the mechanism behind any observed correlations.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Registered Tier

Research Team

Owner:

Zheyang Wu - Late Career Tenured Researcher, Worcester Polytechnic Institute
Aaron Mathieu - Teacher/Instructor/Professor, Acton-Boxborough Regional School District
Yoyo Wu - Student, Acton-Boxborough Regional School District
Nandita Ganesh - Student, Acton-Boxborough Regional School District
Deepika Bhardwaj - Student, Acton-Boxborough Regional School District

GWAS Analysis 4

The purpose of this study is to investigate the genetic underpinnings of Alzheimer’s disease and other dementia types within the dementia cohort of the All of Us dataset. This includes the examination of known genetic loci for deviations linked to…

Scientific Questions Being Studied

The purpose of this study is to investigate the genetic underpinnings of Alzheimer’s disease and other dementia types within the dementia cohort of the All of Us dataset. This includes the examination of known genetic loci for deviations linked to specific Alzheimer’s subtypes and the prediction of the effectiveness of medications based on the presence of specific genetic markers. This field of research is important for developing a better understanding of the genetic factors in dementia, potentially leading to more effective and personalized treatments

Project Purpose(s)

Educational

Scientific Approaches

The intention of this workspace is to utilize the dementia cohort from the All of Us dataset, to predict phenotypic differences dependent on dementia classification type, and allelic data. The analysis is in- tended to involve machine learning tools and statistical techniques such as deep learning models and logistic regression. Clustering algorithms will also be used to identify possible heterogeneity within the dementia cohort.

Anticipated Findings

Anticipated outcomes of this study include the identification of genetic variations associated with spe- cific subtypes of Alzheimer’s disease and insights into how genetic profiles influence the effectiveness of medications in dementia treatment and personalized medicine approaches in treating related conditions. Findings in these areas could open new avenues for research into Alzheimer’s disease and pave the way for more tailored approaches in clinical settings, enhancing the overall management of dementia.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Controlled Tier

Research Team

Owner:

Wesley Lo - Graduate Trainee, Worcester Polytechnic Institute

Collaborators:

Zheyang Wu - Late Career Tenured Researcher, Worcester Polytechnic Institute

ADRD Exploration

Alzheimer’s disease (AD) is a degenerative brain disease which affects about 5.7 million Americans, and the number will grow to about 13.8 million by mid-century. AD and related dementia (ADRD) is a multifactorial and heterogeneous disorder that is driven by…

Scientific Questions Being Studied

Alzheimer’s disease (AD) is a degenerative brain disease which affects about 5.7 million Americans, and the number will grow to about 13.8 million by mid-century. AD and related dementia (ADRD) is a multifactorial and heterogeneous disorder that is driven by a constellation of genetic and environmental risk and protective factors.

Project Purpose(s)

Disease Focused Research (Alzheimer's disease)

Scientific Approaches

Cohort Study and Genetic Analysis

Anticipated Findings

This project will provide practice-based evidence to improve our understanding of the use of selected medications in people with ADRD. And the relationship between ADRD population and gene APOE.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Controlled Tier

Research Team

Owner:

Yu Hou - Early Career Tenure-track Researcher, University of Minnesota

Summary Statistics _ECAD_AD

Exploring the association between carotid atherosclerosis and increased risk of Alzheimer's and Non Alzheimer's Dementia

Scientific Questions Being Studied

Exploring the association between carotid atherosclerosis and increased risk of Alzheimer's and Non Alzheimer's Dementia

Project Purpose(s)

Disease Focused Research (Alzheimer's disease & Extracranial carotid Atherosclerosis)

Scientific Approaches

Summary statistics (mean +- SD) will be presented for demographic characteristics, overall and stratified by ECAD category. Logistic regression analysis will be used to examine the association between Alzheimer's / non-Alzheimer's dementia; we will fit both crude models and models adjusted for covariates (determined based on literature). OR(95% CI) will be reported for all models.

Dataset descriptions: [to be added]

Anticipated Findings

There is growing awareness that extracranial carotid atherosclerosis disease (ECAD) is associated with Alzheimer’s disease and other dementias risk. Despite this, clinical management of ECAD does not involve risk stratification or modification for ADRD. We hope to explore this further.

Demographic Categories of Interest

Data Set Used

Registered Tier

Research Team

Owner:

Cris Escareno - Project Personnel, University of Arizona

Impact of APOE4 on survival from diagnosis of AD dementia in diverse populations

The APOE4 allele is the major susceptibility gene for developing Alzheimer’s disease (AD) at older ages. APOE has 3 common alleles (APOE2, 3 &4), giving rise to 6 genotypes (APOE2/2, 2/3, 2/4, 3/3, 2/4, 3/4 & 4/4). In comparison to…

Scientific Questions Being Studied

The APOE4 allele is the major susceptibility gene for developing Alzheimer’s disease (AD) at older ages. APOE has 3 common alleles (APOE2, 3 &4), giving rise to 6 genotypes (APOE2/2, 2/3, 2/4, 3/3, 2/4, 3/4 & 4/4). In comparison to APOE3/3, the most common genotype, each copy of the APOE4 allele is associated with higher risk of AD dementia & younger median age at dementia onset. The impact of APOE4 on risk, rate of decline, & differential effects of the first AD-modifying disease medications has begun to have a major impact on the fight against AD. Recent studies in relatively small cohorts raised the possibility that APOE4 has a smaller impact on AD risk in African American/Black & Hispanic/Latino than in non-Hispanic persons. Confirming that possibility in large real-world cohort could have major implications for research & care in these underrepresented groups, as well as efforts to discover protective mechanisms that could be targeted by future AD-modifying & prevention therapies.

Project Purpose(s)

Disease Focused Research (Alzheimer’s Disease)

Scientific Approaches

We proposed to capitalize on longitudinal real-world electronic health record (EHR) data from All of Us to characterize differential risk of progressing to clinical diagnosis of probable AD dementia in APOE4 carriers, including homozygote (HM, 4/4), heterozygote (HT, 3/4) & non-carriers (NC, 3/3) in African American/Black, Hispanic/Latino & non-Hispanic participants. Data from participants with these genotypes who are initially ages 60-80, don’t have initial diagnosis of AD dementia & have 5+ years of subsequent EHR data. Survival analyses will control for potential confounds of age, sex, education & if available an indicator of SES. To test our hypothesis with improved statistical power, we will combine HM & HT into an aggregate APOE4 carrier group, compare survival from AD dementia in initial analysis & control for the potential confound of differences among ethnic/racial groups in the carrier group. Exploratory analyses characterize HM vs NC & HT vs NC in the 3 ethnic/racial groups.

Anticipated Findings

We hypothesize that the impact of APOE4 on a person’s AD risk is attenuated in these underrepresented groups (URGs). Confirming that possibility in a large real-world cohort could have major implications for research and care in these URGs, as well as the effort to discover protective mechanisms that could be targeted by future AD-modifying and prevention therapies.

Demographic Categories of Interest

Data Set Used

Controlled Tier

Research Team

Owner:

Valentina Ghisays - Research Associate, Banner Health
Jason Karnes - Early Career Tenure-track Researcher, University of Arizona
Ignazio Piras - Early Career Tenure-track Researcher, Translational Genomics Research Institute
Ehsan Khajouei - Project Personnel, University of Arizona
Dhruman Goradia - Senior Researcher, Banner Health

Searching known Alzheimer's disease causing genes

We will screen known disease causing Alzheimer's disease genes, including APP, PSEN1 and PSEN2. We will create three cohorts including individuals with Alzheimer's disease, dementia and controls with no neurological disease history. We aim to replicate known associations and find…

Scientific Questions Being Studied

We will screen known disease causing Alzheimer's disease genes, including APP, PSEN1 and PSEN2. We will create three cohorts including individuals with Alzheimer's disease, dementia and controls with no neurological disease history. We aim to replicate known associations and find novel variant in these genes.

Project Purpose(s)

Disease Focused Research (Alzheimer's disease, dementia)
Ancestry

Scientific Approaches

We will use annotated VCF files to look for our genes of interest. We will create three cohorts including individuals with Alzheimer's disease, dementia and controls with no neurological disease history. We will examine the pathogenicity of candidate variant using VEP-LOFTEE.

Anticipated Findings

This study will potentially replicate known associations and find novel variant in these genes. Moreover large scale sequencing data will help unravel the frequency of disease causing variants in each of these genes.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Controlled Tier

Research Team

Owner:

Marzieh Khani - Research Fellow, National Institute on Aging (NIH - NIA)

Collaborators:

Fulya Akcimen - Research Fellow, National Institute on Aging (NIH - NIA)

neurological gene screening

We will screen the genes implicated in neurodegenerative diseases. We will create five cohorts, including individuals with Alzheimer's disease, dementia, Parkinson's disease, amyotrophic lateral sclerosis (ALS), and controls with no neurological disease history. We aim to replicate the known associations…

Scientific Questions Being Studied

We will screen the genes implicated in neurodegenerative diseases. We will create five cohorts, including individuals with Alzheimer's disease, dementia, Parkinson's disease, amyotrophic lateral sclerosis (ALS), and controls with no neurological disease history. We aim to replicate the known associations with neurological diseases and identify novel associations.

Project Purpose(s)

Disease Focused Research (neurodegenerative disease)
Ancestry

Scientific Approaches

We will use annotated VCF files to look for our genes of interest. We will create five cohorts, including individuals with Alzheimer's disease, dementia, Parkinson's disease, amyotrophic lateral sclerosis (ALS), and controls with no neurological disease history. We will examine the pathogenicity of candidate variants using VEP-LOFTEE.

Anticipated Findings

This study will potentially replicate known associations and find novel variants in these genes with neurodegenerative diseases. Moreover, large-scale sequencing data will help unravel the frequency of disease-causing variants in each of these genes.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Controlled Tier

Research Team

Owner:

Fulya Akcimen - Research Fellow, National Institute on Aging (NIH - NIA)

Collaborators:

Suleyman Akerman - Research Fellow, Johns Hopkins University

Duplicate of Dementia-Hypertension-Diabetes-2_DatasetV3

Alzheimer’s disease is a neurodegenerative condition characterized by a progressive decline in cognitive function (dementia). Studies suggest that patients with elevated blood pressure (hypertension) are at risk of Alzheimer’s disease type dementias. High blood sugar levels or Type2 Diabetes Mellitus…

Scientific Questions Being Studied

Alzheimer’s disease is a neurodegenerative condition characterized by a progressive decline in cognitive function (dementia). Studies suggest that patients with elevated blood pressure (hypertension) are at risk of Alzheimer’s disease type dementias. High blood sugar levels or Type2 Diabetes Mellitus may also be associated with an increased risk of dementia. Some minority populations may have an increased incidence of hypertension and diabetes. For example, African Americans have a higher incidence of hypertension. Therefore we will to investigate the grouping of racial and ethnic categories, with respect to the incidence of hypertension, diabetes and dementia, to determine whether minority groups have a stronger association between dementia and co-morbidities by race/ ethnicity.
The goal of this demonstration project is to validate previous research showing potential interactions between dementia, diabetes, and hypertension, with an explicit consideration of race/ ethnicity.

Project Purpose(s)

Disease Focused Research (dementia)
Methods Development
Other Purpose (This work is the result of an All of Us Research Program Demonstration Project. Demonstration Projects are efforts by the All of Us Research Program designed to meet the goal of ensuring the quality and utility of the Research Hub as a resource for accelerating precision medicine. This work has been approved, reviewed, and overseen by the All of Us Research Program Science Committee and Data and Research Center to ensure compliance with program policy.)

Scientific Approaches

Data from participants (aged 40 or over) will be subjected to statistical analysis to identify interactions between the incidence of dementia, Diabetes, and Hypertension, and self-identified Race/ Ethnicity. We will only analyze participants in this age group, because the incidence of dementia is very low in patients younger than 40. We will only analyze patients with electronic health care data, because we have to ensure that patients have not had a diagnosis of hypertension, dementia or diabetes.

The statistical analysis package R will be used to create contingency tables, perform chi-squared and Cochran-Mantel-Haenszel tests. Figures will be created in R.

Anticipated Findings

We expect that our data will confirm an increased rate of dementia in African Americans with hypertension and diabetes, compared to white participants. We will determine whether other minorities also see a difference in incidence of dementia, hypertension diabetes and interactions between the them.

If there is an increased incidence of dementia in people with hypertension or diabetes, this may suggest that populations with these disorders need more careful monitoring of their conditions, as they may increase the chance of developing dementia. potentially future All of Us projects may be able to determine whether long term control of hypertension (or Diabetes/ blood glucose) may reduce the potential for developing dementia.

Demographic Categories of Interest

Data Set Used

Registered Tier

Research Team

Owner:

Robert Meller - Mid-career Tenured Researcher, Morehouse School of Medicine

Collaborators:

Shashwat Deepali Nagar - Graduate Trainee, Georgia Institute of Technology
Sonali Gupta - Research Assistant, National Institutes of Health (NIH)
Roxana Loperena Cortes - Other, All of Us Program Operational Use
Francis Ratsimbazafy - Other, All of Us Program Operational Use
King Jordan - Mid-career Tenured Researcher, Georgia Institute of Technology
Kelsey Mayo - Other, All of Us Program Operational Use
Juan Kehoe - Senior Researcher, All of Us Program Operational Use
Elena Moseyko - Project Personnel, All of Us Program Operational Use

Alzheimer's Disease and APOE4

We want to look at Alzheimer's disease and how APOE4 impacts the progression. We also want to do more analysis on the connection between hypertension and Alzheimer's disease.

Scientific Questions Being Studied

We want to look at Alzheimer's disease and how APOE4 impacts the progression. We also want to do more analysis on the connection between hypertension and Alzheimer's disease.

Project Purpose(s)

Disease Focused Research (Alzheimer's disease)

Scientific Approaches

We want to look at genomic data. More specifically we want to look at the APOE4 gene. The datasets we will be using are the alzheimers dataset and also looking at the VCF files.

Anticipated Findings

We want to see if there is a connection between APOE4 and alzheimers. Further analysis, we also want to see if there is any association between hypertension and alzheimers through the APOE4 gene.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Controlled Tier

Research Team

Owner:

Gerardo Rodriguez - Undergraduate Student, Arizona State University

CVD and Alzheimer's in Menopause Transition

The purpose of this research is to investigate the impacts of cardiovascular disease (CVD) on Alzheimer's patients in the menopausal demographic (specifically post menopause). We will explore how external lifestyle factors like diet ad exercise may significantly influence the development…

Scientific Questions Being Studied

The purpose of this research is to investigate the impacts of cardiovascular disease (CVD) on Alzheimer's patients in the menopausal demographic (specifically post menopause). We will explore how external lifestyle factors like diet ad exercise may significantly influence the development of CVD. The study will also formulate and develop effective strategies to mitigate the risk of CVD during menopause transition. Based on the data collected interpretive strategies will be implemented to examine the role of physical activity, especially cardio exercises, in reducing the risk of CVD.

Project Purpose(s)

Population Health
Educational

Scientific Approaches

Patient data will be extracted and analyzed in the All of Us Database to examine the frequency of CVD and Alzheimers amongst women populations undergoing menopause transition in the US. Age, gender, and ethnicity will be used to gather a diverse cohort. If time allows, short read genome sequencing on the APOE gene will be extracted in Alzheimer's patients to outline specific APOE single nucleotide polymorphisms to examine risk of CVD (E2-4 mutations). Estrogen levels will be extracted to plot levels against hypertensive status. The study will focus on patients in the 50+ age range, and a control will be obtained from the 18-35 demographic. Progesterone will also be a variable included in the estrogen levels. Four primary diseases (hypertension, sleep apnea, insomnia, and hypercholesterolemia) and their potential correlation to Alzheimer's and risk for CVD related death in the post menopause demographic.

Anticipated Findings

Based on previous results on the four diseases found the primary incidence rates was found within white females in mid stage of life. Menopausal hormonal shifts were hypothesized to be linked to an increased risk of hypertension, hypercholesterolemia, insomnia, and sleep apnea, which each contributed to an elevated CVD occurrence. Based on previous findings the All of Us database needed more diversity to properly evaluate the spread of risk across the various populations included in the study. The findings will contribute the importance of developing additional diagnostic tests by increasing funding for womens health programs to assist with awareness for the affected demographics. Moreover by increasing funding more research intervention strategies may be developed to provide education on lowering risk of CVD in postmenopausal females.

Demographic Categories of Interest

Data Set Used

Controlled Tier

Research Team

Owner:

Augustus Storm - Undergraduate Student, Arizona State University

Dimentia GWAS Analysis

The purpose of this study is to investigate the genetic underpinnings of Alzheimer’s disease and other dementia types within the dementia cohort of the All of Us dataset. This includes the examination of known genetic loci for deviations linked to…

Scientific Questions Being Studied

The purpose of this study is to investigate the genetic underpinnings of Alzheimer’s disease and other
dementia types within the dementia cohort of the All of Us dataset. This includes the examination
of known genetic loci for deviations linked to specific Alzheimer’s subtypes and the prediction of the
effectiveness of medications based on the presence of specific genetic markers. This field of research is
important for developing a better understanding of the genetic factors in dementia, potentially leading
to more effective and personalized treatments

Project Purpose(s)

Educational

Scientific Approaches

The intention of this workspace is to utilize the dementia cohort from the All of Us dataset, to predict
phenotypic differences dependent on dementia classification type, and allelic data. The analysis is in-
tended to involve machine learning tools and statistical techniques such as deep learning models and
logistic regression. Clustering algorithms will also be used to identify possible heterogeneity within the
dementia cohort.

Anticipated Findings

Anticipated outcomes of this study include the identification of genetic variations associated with spe-
cific subtypes of Alzheimer’s disease and insights into how genetic profiles influence the effectiveness of
medications in dementia treatment and personalized medicine approaches in treating related conditions.
Findings in these areas could open new avenues for research into Alzheimer’s disease and pave the way
for more tailored approaches in clinical settings, enhancing the overall management of dementia.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Registered Tier

Research Team

Owner:

Wesley Lo - Graduate Trainee, Worcester Polytechnic Institute

GWAS Dementia 2

The purpose of this study is to investigate the genetic underpinnings of Alzheimer’s disease and other dementia types within the dementia cohort of the All of Us dataset. This includes the examination of known genetic loci for deviations linked to…

Scientific Questions Being Studied

The purpose of this study is to investigate the genetic underpinnings of Alzheimer’s disease and other dementia types within the dementia cohort of the All of Us dataset. This includes the examination of known genetic loci for deviations linked to specific Alzheimer’s subtypes and the prediction of the effectiveness of medications based on the presence of specific genetic markers. This field of research is important for developing a better understanding of the genetic factors in dementia, potentially leading to more effective and personalized treatments

Project Purpose(s)

Educational

Scientific Approaches

The intention of this workspace is to utilize the dementia cohort from the All of Us dataset, to predict phenotypic differences dependent on dementia classification type, and allelic data. The analysis is in- tended to involve machine learning tools and statistical techniques such as deep learning models and logistic regression. Clustering algorithms will also be used to identify possible heterogeneity within the dementia cohort.

Anticipated Findings

Anticipated outcomes of this study include the identification of genetic variations associated with spe- cific subtypes of Alzheimer’s disease and insights into how genetic profiles influence the effectiveness of medications in dementia treatment and personalized medicine approaches in treating related conditions. Findings in these areas could open new avenues for research into Alzheimer’s disease and pave the way for more tailored approaches in clinical settings, enhancing the overall management of dementia.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Controlled Tier

Research Team

Owner:

Wesley Lo - Graduate Trainee, Worcester Polytechnic Institute

Collaborators:

Zheyang Wu - Late Career Tenured Researcher, Worcester Polytechnic Institute

presence of Alzheimer's variants within diverse populations

In the United the States, the Hispanic population is a unique diverse population with a diverse genetic makeup. The prevalence of Alzheimer's Disease has been reported to be increasing in the Hispanic population. Are there variants that present with higher…

Scientific Questions Being Studied

In the United the States, the Hispanic population is a unique diverse population with a diverse genetic makeup. The prevalence of Alzheimer's Disease has been reported to be increasing in the Hispanic population. Are there variants that present with higher frequency within Hispanic populations when compared to other Non-Hispanic populations.

Project Purpose(s)

Disease Focused Research (Alzheimer's disease)
Population Health
Educational
Ancestry
Other Purpose (All of Us research scholar project)

Scientific Approaches

I would like to explore the data within the workbench, in order to learn and explore variants linked to Alzheimer,s Disease, potentially looking at comparative analysis and descriptive methods.

Anticipated Findings

The identification of lesser-known variants related to Alzheimer's disease can build the scientific knowledge base In better understanding and exploring the presence of multiple variants linked to Alzheimer's Disease, within Hispanic populations.

Demographic Categories of Interest

Data Set Used

Controlled Tier

Research Team

Owner:

carolina frantz - Graduate Trainee, Duquesne University

hypertensionDementia

Objective: This study aims to quantify the link between antihypertensive therapies and Alzheimer's disease risk, expanding on the established relationship between primary hypertension and AD. Introduction: Previous research has identified primary hypertension as a key risk factor for AD. Effective…

Scientific Questions Being Studied

Objective: This study aims to quantify the link between antihypertensive therapies and Alzheimer's disease risk, expanding on the established relationship between primary hypertension and AD.

Introduction: Previous research has identified primary hypertension as a key risk factor for AD. Effective hypertension management may impact AD progression and cognitive decline.

Methods: Leveraging All of Us data, we'll longitudinally analyze individuals, stratifying by hypertension status. Those with primary hypertension will be categorized by antihypertensive drug class: Thiazide-type diuretics, Calcium channel blockers, ACE inhibitors, and ARBs. We'll assess cognitive performance trajectories, adjusting for confounders, precision variables, and time-varying variables influenced by prior interventions. The confounders we will incorporate into our models will be derived by searching a biomedical knowledge graph derived from both the literature and biomedical ontologies.

Project Purpose(s)

Disease Focused Research (Alzheimer's disease)
Population Health
Social / Behavioral
Drug Development
Methods Development
Control Set

Scientific Approaches

Research: Our inaugural All of Us study aims to conduct a scientifically rigorous retrospective case-control investigation. We've devised techniques to identify confounding variables for model integration.

Data: We intend to employ All of Us participant health data to explore the impact of hypertension treatments on AD risk for individuals aged 65+ with a 10+ year history. We'll consider:

Exposure: Hypertension treatments by drug class for new users, mitigating "time-zero bias" as much as possible.
Outcome: Dementia presence (AD, VaD, mixed, cerebrovascular dementia) using ICD-9//10 codes and Memantine, donepezil prescriptions.
Time-varying confounders: Cognitive performance, physical activity. Additional covariates: Age, sex, race/ethnicity, APOe2/3/4 status, social determinants of health, vascular comorbidities (stroke, heart attack), and other research factors (sleep apnea, vitamin D deficiency, COPD).

Tools: We will use marginal structural models for longitudinal data analysis.

Anticipated Findings

Conclusion: This comprehensive study aims to elucidate the differential effects of specific antihypertensive medication classes on the risk and progression dynamics of Alzheimer's disease and affiliated dementias. The outcomes promise to enhance our understanding of the intricate nexus between hypertension management and the evolution of dementia-related outcomes.

Demographic Categories of Interest

Data Set Used

Registered Tier

Research Team

Owner:

Scott Malec - Early Career Tenure-track Researcher, University of New Mexico and University of New Mexico Health Sciences Center

Duplicate of Precision and Diversity in Alzheimer's Disease and Dementia

Alzheimer's Disease (AD) is a progressive neurological disorder and the most common cause of dementia. Genetic research has greatly increased our knowledge for the genetic basis of this disease, however despite this, we have made little headway in the discovery…

Scientific Questions Being Studied

Alzheimer's Disease (AD) is a progressive neurological disorder and the most common cause of dementia. Genetic research has greatly increased our knowledge for the genetic basis of this disease, however despite this, we have made little headway in the discovery of disease modifying or preventative therapeutics. Additionally, genetic research for AD, as well as other diseases, is dominated by those of European ancestry. To address these issues, we are focussing on 2 scientific questions.

1. Are there genetic modifiers that affect risk for developing AD in carriers of the APOE e4 genotype? Having just one copy of the APOE e4 genotype greatly increases the risk for developing AD, however, many people possess two copies of the APOE e4 genotype and never develop AD.

2. What variants affect AD risk for non-European populations? Leveraging different LD patterns, we can increase the power to detect additional risk variants.

Project Purpose(s)

Disease Focused Research (Alzheimer's Disease and Dementia)

Scientific Approaches

Our plan to address these 2 questions is through a series of meta-analyses and follow-up work. The plan for question 1 is to perform a Genome-Wide Association Study of AD in APOE e4 carriers only. We will use the AllofUs genetic datasets as well as the demographics age, sex, and AD disease status. We will use Plink software to carry out this analysis. We will then meta-analyze these results with summary statistics from other AD cohorts. Follow-up will include FUMA and gene burden tests to further investigate any potential findings. The plan for question 2 is to perform a Trans-Ancestry Meta Analysis to nominate new risk variants. We will use Plink to perform GWAS is populations that have been defined by genetic ancestry, and then will use MR-MEGA software to meta-analyze the summary statistics in order to nominate new risk variants that may have different effects on disease risk in different ancestry populations.

Anticipated Findings

The anticipated findings are two-fold: the first is to nominate genetic variants that effect AD risk in APOE e4 carriers. This is one of many important steps on the journey to precision medicine. It is unlikely that we will find a therapeutic that works for all types of AD and dementia. Instead, we must focus on specific types of disease. By discovering additional genetic modifiers of risk (why do some APOE e4 carriers not get AD?) we can potentially nominate new targets that have a better chance at disease modification. Second, we need to make an active effort as a research community to include non-European populations in genetic research. Much less is known about the genetic basis of AD in non-European populations. We aim to nominate new AD risk variants that are present in multiple and different populations so that we can increase our knowledge and treatment of those populations.

Demographic Categories of Interest

Data Set Used

Controlled Tier

Research Team

Owner:

Hampton Leonard - Graduate Trainee, National Institute on Aging (NIH - NIA)

Impact of sex at birth and treatment pathway on Alzheimer's Disease

How effective is treatment of Alzheimers disease (AD) using donepezil, memantine, or both in combination based on the AllofUs dataset? Additionally, how often each of these different treatments administered to the general patient population in terms of sex at birth?…

Scientific Questions Being Studied

How effective is treatment of Alzheimers disease (AD) using donepezil, memantine, or both in combination based on the AllofUs dataset? Additionally, how often each of these different treatments administered to the general patient population in terms of sex at birth?

Through answering these questions, more data will be contributed to the existing literature on the effectiveness of these three therapies on a more diverse patient population. Additionally, the relationship between sex at birth and treatment administered to effectively treat the symptoms of AD will be analyzed. This will help determine if there is biased treatment administration based on sex at birth, and could help determine whether a treatment path seems to treat symptoms better depending on sex at birth.

Project Purpose(s)

Disease Focused Research (Alzheimer's disease)
Educational

Scientific Approaches

I plan to mainly use the "Conditions" dataset, the "Drug exposure" dataset. Using the AllofUs coding tools, I hope to compare all three drug combinations and how they were administered to participants of different sexes. Using an umbrella study design, I will look at the three different drug treatment options and attempt to determine a connection between the symptoms described by participants and their treatment course as well as how this relates to their sex at birth.

Anticipated Findings

I anticipate finding that combination therapy is the most effecting treatment for slowing the progression of Alzheimer's disease (based on the literature) and that this therapy is most effective in patients born female at birth. This is the expected outcome because of the literature recently researched, and because Alzheimer's disease occurs more often in females. Researching these questions in the AllofUs, data set will be contributed to the scientific knowledge by showing how these therapies impact a more diverse cohort of patients while also looking into the impact of sex at birth on Alzheimer's disease progression when using 3 different treatment pathways.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Registered Tier

Research Team

Owner:

Jacob Seeker - Graduate Trainee, Northeastern University

Collaborators:

Michael Wilczek - Research Fellow, Northeastern University

Alz + Blood Transfusion

I am interested in looking at any positive correlation between a history of blood transfusion + later development of Alzheimer's disease; I am working w/ the UKB database to compare data results for validation purposes and potential to correlate to…

Scientific Questions Being Studied

I am interested in looking at any positive correlation between a history of blood transfusion + later development of Alzheimer's disease; I am working w/ the UKB database to compare data results for validation purposes and potential to correlate to imaging findings. The goal is to identify if there is a potential risk of blood-borne transmission of amyloid-like pathologies, like Alzheimer's. This is a question I developed following a recent JAMA article looking at the correlation between the risk of spontaneous ICHin blood transfusion recipients who received blood from later identified CAA+ patients (something that would not have been known at the time of donation). There is an increased risk of ICH and possible transmission of CAA via blood - given this relationship its possible amyloid-like pathology follows a similar pattern of previously unidentified transmission, especially given that patients who donate may not realize that they later on will develop Alzheimers.

Project Purpose(s)

Disease Focused Research (Alzheimer's disease)
Population Health
Control Set

Scientific Approaches

Datasets will include patients with a history of transfusion w/ and w/o the development of Alzheimer's, and patients with Alzheimer's w/ and w/o a history of both blood donation or transfusion recipients. Then we will analyze for any statistically significant correlation. If correlation is demonstrated we plan to validate against the UKB dataset, if validated we will correlate findings with imaging and polygenic risk stratification data. The big question is more - is there a correlation, if not then we may not move forward with further analysis as it likely wouldn't be indicated.

Anticipated Findings

My hypothesis is that there may be some positive correlation. Previous studies have demonstrated that there is not an identified (as of yet) transmissible element in blood transfusions for amyloid-B pathology, however, those studies have specifically looked for an identifiable element within patients' blood, e.g. some kind of biomarker of Alzhimer's that may indicate a patient's propensity for later development (like measuring high blood pressure as a correlate for increased cardiovascular disease risk). What has not been looked at is the rate or potential relationship between blood donation in presumably healthy patients who later develop Alzheimer's and the risk that their transfused recipients later develop Alzheimer's - this is largely because these diseases are slow progressing, and identification of amyloid-like neuropathology is usually post-mortem or after severe cognitive decline, meaning honestly the data to correlate is emerging.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Registered Tier

Research Team

Owner:

Amanda Watters - Graduate Trainee, Tulane University

AD project

I'm interested in factors that trigger Alzheimer's Disease, which has been found a efficacious treatment.

Scientific Questions Being Studied

I'm interested in factors that trigger Alzheimer's Disease, which has been found a efficacious treatment.

Project Purpose(s)

Educational

Scientific Approaches

I'm planning to use multivariable regression analysis to explore factors that might trigger Alzheimer's Disease.

Anticipated Findings

I'm expecting to see an association between environmental factors and Alzheimer's Disease. These findings will contribute to the prediction and prevention of Alzheimer's Disease.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Controlled Tier

Research Team

Owner:

Xinhui Yang - Graduate Trainee, Louisiana State University Health Sciences Center, New Orleans

Analyze time-varying genetic effect with the integration of EHR

We will integrate genetic data with the electronic health records to analyze the underlying disease pathways. Specifically, we propose to investigate the time-varying effect genetic effect. The specific questions we will ask are: 1. How does the time-varying genetic effect…

Scientific Questions Being Studied

We will integrate genetic data with the electronic health records to analyze the underlying disease pathways. Specifically, we propose to investigate the time-varying effect genetic effect. The specific questions we will ask are:
1. How does the time-varying genetic effect play a role in late onset Alzheimer’s diseases?
2. Can we develop a framework to monitoring the potential sign of the late onset Alzheimer’s diseases before it is diagnosed?
3. Integrated with other lab measurements and brain image resources, can we utilize deep learning to develop a unified system for the detection and prediction of Alzheimer’s diseases?
This study will focus on the statistical methodology development and applications using All of Us data. We anticipate that the proposed investigations will contribute to novel statistical advances and a better understanding of the genetic structures associated with Alzheimer’s diseases.

Project Purpose(s)

Disease Focused Research (Alzheimer's disease)
Methods Development
Ancestry

Scientific Approaches

We will develop innovative statistical methods based on quantile regressions construct a reference system to monitoring the development of Alzheimer’s diseases. An intuitive idea would be growth chart to monitor the height and weight of a child and give the percentile of him/her in the cohort. By integrating genetic data with electronic health records, such as brain images and recognition tests, we will be able to build a more comprehensive system to describe the patient’s disease progression given the cohort as a reference. This will also help prevent or warn the appearance of the Alzheimer’s at an early stage. Technically speaking, we will use low-rank approximation, data integration, and quantile regression to achieve this goal. We will develop statistical estimation and inference tools to quantify the uncertainty, and illustrate the methods by applications in All of Us data.

Anticipated Findings

For this study, we expect to have contributions in statistical methodology and genetics. From a statistical perspective, integrating multi-source data is always a challenging task, due to the errors in the observed data and the heterogeneity caused by unmeasured confounders. We will use quantile regression, a robust way of modeling, to reduce the effect from data heterogeneity, and low-rank approximation, to extract the main feature from different types of the data (e.g., brain images, lab measurements). From a genetic perspective, it is novel and in need to investigating time-varying genetic effect, especially for aging-related diseases. Due to the increasing amount of patients suffered from Alzheimer’s diseases, analyzing the time-varying genetic effect, along with individual-specific EHR data, will greatly unfold the patient’s specific disease progressions and lead to more precise treatment.

Demographic Categories of Interest

Data Set Used

Controlled Tier

Research Team

Owner:

Tianying Wang - Early Career Tenure-track Researcher, Colorado State University

lab measurements of Alzheimer's

We use lab measurements of people with Alzheimer's condition and do correlation analysis of it. Emphasizing correlation does not mean causation, we try to check if common missing value imputation methods can be used.

Scientific Questions Being Studied

We use lab measurements of people with Alzheimer's condition and do correlation analysis of it. Emphasizing correlation does not mean causation, we try to check if common missing value imputation methods can be used.

Project Purpose(s)

Educational

Scientific Approaches

Dataset with lab measurements (numerical features). Artificial missingness is generated using Missing at Random generator. Then MICE imputation algorithm is used and check how well it was able to impute using mean square error (MSE) metric. We repeat the whole process 10 times to get the average MSE score.

Anticipated Findings

Effectiveness of MICE imputation on EHR dataset particularly on Alzheimer's patients

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Registered Tier

Research Team

Owner:

Shourav Rabbani - Graduate Trainee, Tennessee State University

Identifying Patients with Alzheimer's Disease

This study seeks to address several critical scientific questions with direct relevance to Alzheimer's disease and public health. By exploring patients with Alzheimer's and examining their lifestyle choices through survey questionnaires, we aim to uncover insights into the lifestyle factors…

Scientific Questions Being Studied

This study seeks to address several critical scientific questions with direct relevance to Alzheimer's disease and public health. By exploring patients with Alzheimer's and examining their lifestyle choices through survey questionnaires, we aim to uncover insights into the lifestyle factors associated with cognitive preservation or decline, the potential modifiable factors for preventing Alzheimer's, and how personalized care plans can be tailored to the unique preferences of Alzheimer's patients. These questions are pivotal as they hold the potential to advance our understanding of the disease, inform public health initiatives, improve the lives of Alzheimer's patients, and guide evidence-based care and preventive strategies.

Project Purpose(s)

Social / Behavioral

Scientific Approaches

For this study, I plan to employ a multifaceted scientific approach that combines exploratory data analysis (EDA) with quantitative and qualitative research methods. To begin, EDA will serve as the cornerstone, allowing me to gain valuable insights and uncover hidden patterns within the datasets collected through survey questionnaires. These datasets will encompass a wide range of lifestyle factors, including dietary habits, physical activity, cognitive engagement, and social interactions among individuals with Alzheimer's. Qualitative research methods, such as interviews or open-ended survey questions, will complement the quantitative analyses, enabling a deeper understanding of the personal experiences and perspectives of individuals living with Alzheimer's. This integrated approach will facilitate a holistic examination of the research question and yield comprehensive and well-informed conclusions.

Anticipated Findings

The anticipated findings from this study hold the potential to make significant contributions to the field of Alzheimer's disease and gerontology. By investigating the specific lifestyle choices and trends among individuals living with Alzheimer's, we aim to uncover valuable insights, including the identification of beneficial lifestyle factors, insights into disease progression, potential preventative measures, personalized care development, and relevance to healthcare policy. These findings are not only focused on enhancing the lives of people with Alzheimer's but also on advancing the broader scientific understanding of how lifestyle choices impact the progression and management of the disease. Through this research, we hope to make a meaningful difference in the lives of those affected by Alzheimer's and improve the strategies for care, support, and prevention in the field.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Registered Tier

Research Team

Owner:

Matthew Harris - Undergraduate Student, Tennessee State University

Hackathon 2023

For the upcoming NIA Hackathon, we will use AllOfUs as an exercise in querying AllofUs database, data exploration/summary stats, visualizing association between environmental factors and genetic factors (polygenic risk score) for Alzheimer’s & Parkinson’s disease, survival analysis. Specifically, the hackathon…

Scientific Questions Being Studied

For the upcoming NIA Hackathon, we will use AllOfUs as an exercise in querying AllofUs database, data exploration/summary stats, visualizing association between environmental factors and genetic factors (polygenic risk score) for Alzheimer’s & Parkinson’s disease, survival analysis. Specifically, the hackathon group will look at environmental/lifestyle factors: deprivation index, alcohol, smoking, exercise. The goal is to introduce this platform to more NIA teams and encourage use for research purposes.

Project Purpose(s)

Educational

Scientific Approaches

- Generate NDD (AD/PD) cohort and creating dataset w/ covariates
- Descriptive statistics
- Compute polygenic risk scores
- Survival analysis

Anticipated Findings

Hopefully, this exercise will lead to future research collaborations between CARD and NIA LNG in Baltimore.

Demographic Categories of Interest

Data Set Used

Controlled Tier

Research Team

Owner:

Lietsel Jones - Project Personnel, National Institute on Aging (NIH - NIA)
Hampton Leonard - Graduate Trainee, National Institute on Aging (NIH - NIA)

Collaborators:

Peter Wild Crea - Graduate Trainee, National Institute on Aging (NIH - NIA)
Toshiko Tanaka - Research Associate, National Institute on Aging (NIH - NIA)
Richard Oppong - Research Fellow, National Institute on Aging (NIH - NIA)
Abigail Miano-Burkhardt - Research Assistant, National Institute on Aging (NIH - NIA)
Fulya Akcimen - Research Fellow, National Institute on Aging (NIH - NIA)
Jun Ding - Senior Researcher, National Institute on Aging (NIH - NIA)
Sara Bandres Ciga - Other, National Institute on Aging (NIH - NIA)
Chelsea Alvarado - Other, National Institute on Aging (NIH - NIA)

Sleeping Disease & Alzheimer's Disease

Alzheimer's Disease is the most common form of dementia in the United States. This study means to identify phenotypic associations between AD & SD using machine learning (ML). The objective is to identify risk factors of mid-age and aging patients…

Scientific Questions Being Studied

Alzheimer's Disease is the most common form of dementia in the United States. This study means to identify phenotypic associations between AD & SD using machine learning (ML). The objective is to identify risk factors of mid-age and aging patients with SD to develop AD.

Project Purpose(s)

Disease Focused Research (Alzheimer's Disease)

Scientific Approaches

We will use patient data within the observation window to build prediction model. We plan to use ML models including logistic regression, decision tree, and the advanced tree-based model XGboost. Data will be split into training and testing sets. Model hyperparameters will be tuned using a cross validation strategy in the training set.

Anticipated Findings

Based on the proposed study design, we expect to identify the specific phenotypic associations between Sleep Disorders (SD) and Alzheimer's Disease (AD) through the use of Machine Learning (ML) techniques. This will involve identifying various risk factors associated with mid-aged and older patients with SD that may predispose them to developing AD. Our models should illuminate individual-level outcomes for patients with AD, drawing on a range of structured patient records including demographic details, symptom profiles, comorbid conditions, medications, and Social Determinants of Health (SDoH). The outcomes from this study would not only deepen our understanding of the association between SD and AD but also provide practical tools and insights to help mitigate the impact of Alzheimer's disease.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Data Set Used

Registered Tier

Research Team

Owner:

Chang Su - Early Career Tenure-track Researcher, Temple University

Subscribe to Email Updates