Featured workspaces provide examples of data querying, wrangling, and analysis to support you when using the All of Us Researcher Workbench. There are four types of featured workspaces: tutorial workspaces, demonstration projects, phenotype library, and community workspaces.
To access featured workspaces
- Log in to the Researcher Workbench.
- Click “
.”
- Click “Featured Workspaces.”
- Click the type of featured workspace: “Tutorial Workspaces,” “Demonstration Projects,” “Phenotype Library,” or “Community Workspaces” for a list of relevant workspaces.
-
Click the title of the workspace you want to open.
Note: Featured workspaces are read only. To edit a featured workspace, you need to duplicate the workspace.
To duplicate a featured workspace
Featured workspaces are READ ONLY unless you duplicate the workspace for your own use.
- Click “
.”
- Click “Duplicate.”
Note: When you duplicate a featured workspace, a copy of the original workspace description is included in the duplicated workspace.
- Edit and/or update the workspace description as necessary.
Note: Read about writing your workspace description for information about each section you see on the screen.
- Answer “yes” or “no” to a review of your research purpose description.
Note: For featured workspaces, you do not need review as the research purpose was written by the original Researcher Workbench user.
- Click “Duplicate Workspace.”
Tutorial workspaces
From basic data manipulation to analysis techniques specific to the All of Us data, tutorial workspaces are a great place to start for learning how to analyze data within the All of Us Researcher Workbench.
Beginner intro to All of Us data and the All of Us Researcher Workbench
The “Beginner Intro to All of Us data and the All of Us Researcher Workbench” featured workspace contains multiple notebooks to help you assess your understanding of the Observational Medical Outcomes Partnership (OMOP), the data structure and organization of the All of Us Research Program, and the programming languages Python, R, and SQL. The notebooks provide the basics you need to know to start and complete your research in the Researcher Workbench.
Getting started with the Registered Tier data
The “Getting started with the Registered Tier data” featured workspace provides an overview of what data are available in the Registered Tier of the current Curated Data Repository. The workspace provides information on how to retrieve electronic health records (EHRs), physical measurements, and survey data. The notebooks provide example code and walk through the data wrangling and analysis process.
Getting started with the Controlled Tier data
The “Getting started with the Controlled Tier data” featured workspace provides an overview of what data are available in the Controlled Tier of the current Curated Data Repository. The workspace does not include an overview of the data that are in both the Registered Tier and the Controlled Tier. The workspace provides information on how to retrieve genomic data. Note: For the Registered Tier data, use the Getting started with the Registered Tier data featured workspace.
Data wrangling in the All of Us Researcher Workbench
The “Data wrangling in the All of Us Researcher Workbench” featured workspace provides new users a basic tutorial to data wrangling in the Researcher Workbench. The workspace provides a step-by-step walkthrough on building cohorts, pulling specific data types, merging and combining data into one dataset, visualizing the data, and analyzing the data with common statistical tests.
Working with All of Us survey data
The “Working with All of Us survey data” featured workspace helps you become familiar with how to query participant provided information (PPI) from survey questions. The notebooks provide example code and walk through the process of extracting and visualizing survey data.
The “Working with All of Us COVID-19 Participant Experience (COPE) survey data” featured workspace helps you become familiar with survey data from the COPE survey. The notebooks provide example code and walk through the process of extracting and visualizing COPE survey data.
The “Working with All of Us physical measurements data” featured workspace helps you become familiar with physical measurements data, which include height, weight, body mass index (BMI), waist circumference, hip circumference, pregnancy status, blood pressure, heart rate, and wheelchair use. The notebooks provide example code and walk through the process of extracting and visualizing physical measurements data.
The “Working with All of Us wearables data” featured workspace helps you become familiar with wearables data. The notebooks provide example code and walk through the process of extracting and visualizing wearables data.
The “Backing up notebooks and intermediate results” featured workspace gives an overview on creating snapshots of notebooks and backups of intermediate results stored in other files such as plot images and derived data. The workspace includes steps for saving snapshots for later review, allowing users to track changes to results in notebooks over time, and creating files such as image files of plots of CSVs of intermediate results.
The “Running notebooks in the background” featured workspace provides a notebook for running code in the background. Note: For your analysis, the cluster will auto-pause after 24 hours. To prevent your cluster from shutting down if your background job takes longer than 24 hours, log in and start any notebook in the workspace to reset the auto-pause timer.
The “Working with All of Us genomic data (Hail-PLINK)” featured workspace helps you get started with the All of Us genomic data and tools. The workspace includes notebooks demonstrating analysis using Hail and PLINK to perform genome-wide association studies (GWAS) using the All of Us genomic data and phenotypic data.
The “Working with All of Us genomic data (CRAM_processing and IGV)” featured workspace teaches you how to copy or localize All of Us CRAM files to your workspace bucket and active cloud environment to look at the file contents with the Integrated Genome Viewer (IGV).
The “Running Workflow Description Language (WDL) with Cromwell” featured workspace explains how to set up Cromwell, how to use the automatically created Cromwell configuration file, and how to write WDL script with the All of Us genomic data as an input. The workspace walks through setting up your cloud environment and using Cromwell to execute an example script, validate_vcf.wdl. The workflow uses the GATK ValidateVariants tool to validate variant call format (VC) files. VCF files, corresponding index files, and the human reference genome assembly are provided to the workflow as inputs.
The “Using Nextflow in the All of Us Researcher Workbench” featured workspace explains how to set up Nextflow, how to use the automatically created Nextflow configuration file, and how to write a Nextflow script with the All of Us genomic data as an input. The workspace walks through setting up your cloud environment and using Nextflow script with the array single sample variant call format (VCF) files as inputs.
The “Using dsub in the All of Us Researcher Workbench” featured workspace explains how to set up dsub, how to run a single job, how to check job status, how to debug a failed job, and how to run the wc command in a parallel manner.
The “Genomics Undergrad Lesson Plan Exemplar” featured workspace provides content for helping mentor students and trainees on projects in the Controlled Tier. The workspace provides a lesson plan to support mentors with onboarding resources, sample workspaces using the programming language R, and more.
The “Reproducing the All of Us SARS-CoV-2 Antibody Study (version 4)” featured workspace provides an overview of the published research, “Antibodies to Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) in All of Us Research Program Participants, 2 January to 18 March 2020.” The workspace utilizes the Curated Data Repository (CDR) version 4 and includes directions for reproducing the results of the study.
The “Reproducing the All of Us SARS-CoV-2 Antibody Study (version 5)” featured workspace provides an overview of the published research, “Antibodies to Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) in All of Us Research Program Participants, 2 January to 18 March 2020.” The workspace utilizes the Curated Data Repository (CDR) version 5 and includes directions for reproducing the results of the study.
The “Working with All of Us long-read data” featured workspace teaches you how to copy or localize All of Us long-read BAM files individually and how to render the Integrated Genome Viewer (IGV) to explore the BAM files. The workspace includes a notebook dedicated to various ways of analyzing BAM localization and a notebook for looking at BAM files in your current environment with IGV.
The “Intro to All of Us genomic data” featured workspace provides an introduction to analyzing All of Us genomic data. The workspace includes notebooks and examples for hands-on experience with using the genomic data to run a genome-wide association study (GWAS) using Hail.
The “Intro to All of Us phenotypic and survey data” featured workspace is intended to familiarize researchers with the survey, electronic health record (EHR) and Fitbit data on the Researcher Workbench. By running the exercises in this workspace, researchers will have a better understanding of how to build a cohort of participants with these data types.
The “Querying All by All results and analysis details” featured workspace demonstrates methods to effectively filter and export summary statistics of interest from the All by All tables.
The “All by All: Curation of drug phenotypes” featured workspace provides additional information about the drug exposure phenotypes for the All by All tables.
The “All by All: Curation of lab measurements phenotypes” featured workspace provides additional information about the lab measurements phenotypes for the All by All tables.
The “All by All: Curation of Personal and Family Health History survey phenotypes” featured workspace provides additional information about the Personal and Family Health History survey phenotypes for the All by All tables.
The “All by All: Curation of phecode phenotypes” featured workspace provides additional information about the phecode phenotypes for the All by All tables.
The “All by All: Curation of phecodeX phenotypes” featured workspace provides additional information about the phecode and phecodeX phenotypes for the All by All tables.
The “All by All: Curation of physical measurements phenotypes” featured workspace provides additional information about the physical measurements phenotypes for the All by All tables.
The “Best practices for All of Us data science” featured workspace demonstrates best practices to query All of Us data and work with environment variables based on frequently asked questions by All of Us Researcher Workbench users during Office Hours or via Help Desk tickets.
The “SAS 101 data fundamentals” featured workspace demonstrates best practices for working with the All of Us data using SAS Studio. The workspace includes how to perform common SAS procedures and how to explore the dataset using SAS.
Demonstration projects
Demonstration projects showcase the quality, utility, and diversity of the All of Us data by replicating end-to-end analyses in previously published studies in the All of Us Researcher Workbench.
Data Quality Report (version 6)
The “Data Quality Report (version 6)” demonstration project provides detailed demographic information about the All of Us participants in the Curated Data Repository (CDR) version 6 (2022Q2R2), as well as a summary of participants by data type. The project includes multiple notebooks: summary of participants by data type, demographic characteristics of participants by data types, and underrepresented in biomedical research breakdown.
Data Quality Report (version 7)
The “Data Quality Report (version 7)” demonstration project provides detailed demographic information about the All of Us participants in the Curated Data Repository (CDR) version 7 (C2022Q4R9), as well as a summary of participants by data type. The project includes multiple notebooks: summary of participants by data type, demographic characteristics of participants by data types, underrepresented in biomedical research breakdown, underrepresented in biomedical research by data type, and genomics by data types.
N3C machine learning PASC/long COVID phenotype algorithm in the All of Us dataset
The “N3C machine learning PASC/long COVID phenotype algorithm in the All of Us dataset” demonstration project is a collaboration between the All of Us Research Program, National COVID Cohort Collaborative (N3C), PCORnet, and NIH/RECOVER to examine and identify participant risk of Long COVID utilizing the N3C's machine learning PASC/Long COVID Phenotype algorithm within the All of Us Researcher Workbench. The XGBoost machine learning model was used to identify potential patients with PASC/Long COVID, which was initially published by Emily et al. These models were subsequently implemented within the All of Us Controlled Tier.
Wearables and the human phenome
The “Wearables and the human phenome” demonstration project examines the associations between physical activity over time measured using participant wearables and incident chronic diseases as determined by electronic health record (EHR) data.
PheWAS smoking
The “PheWAS smoking” demonstration project contains the results of Phenome-Wide Association Studies (PheWAS) to show how the various sources of data contained within the All of Us dataset can be used to inform scientific discovery. The project demonstrates how to implement a PheWAS within the All of Us Researcher Workbench, how to use heterogeneous data sources within the All of Us dataset, and how to develop plots that compare the results of electronic health records (EHR) smoking with participant provided information (PPI) ever smoking and PPI smoking every day PheWAS routines.
The “Cardiovascular risk scoring” demonstration project uses the American Heart Association algorithm/equation to calculate cardiovascular risk scores. The project demonstrates the usage of smoking and race data collected by the All of Us Research Program.
The “Medication sequencing” demonstration project uses medication sequencing developed at Columbia University and the Observational Health Data Sciences and Informatics (OHDSI) program as a way to characterize treatment pathways at scale. The project demonstrates the implementation of the medication sequencing algorithms in the All of Us dataset to show how the various sources of data contained within the All of Us Research Program can be used to characterize treatment pathways at scale.
The “All of Us descriptive statistics” demonstration project applies data visualization libraries to aggregate information about All of Us participants. The project measures age by using the age reflected when the Curated Data Repository (CDR) was generated. The project aims to describe an overview of data types included in beta release CDR; describe participants by age, race, and ethnicity using all the available data types; and describe the underrepresented biomedical research population within the All of Us Research Program participants.
The “Family history in electronic health record (EHR) and participant provided information (PPI)” demonstration project summarizes structured data elements available in the All of Us Registered Tier and compares published survey results to describe data for reuse in disease specific outcomes.
The “Hypertensive prevalence” demonstration project aims to use published methods to replicate known differences in hypertension prevalence in underrepresented biomedical research populations and to illustrate variation in hypertension prevalence in geographic regions of the U.S.
The “Assessment of pathogenic variants across the All of Us Research Program” demonstration project assesses relative frequency of positive findings across the All of Us dataset and compares the aggregate findings with those from other cohorts such as GnomAD. The workspace is associated with the published work by Venner, E., Patterson, K., Kalra, D. et al. titled, “The frequency of pathogenic variation in the All of Us cohort reveals ancestry-driven disparities."
The “Systemic disease and glaucoma” demonstration project aims to externally validate a single-center model’s performance with All of Us data, to develop models trained by the All of Us data and compare their performance to a single-center model, and to share insights about using All of Us data and the All of Us Researcher Workbench with ophthalmology researchers who may be interested in using this data source.
The “Siloed analysis of All of Us and UK Biobank genomic data” demonstration project demonstrates the potential of the All of Us Researcher Workbench for pooled analyses of All of Us and UK Biobank data. The project aims to develop and describe an approved, secure path for connecting UK Biobank data to the All of Us Researcher Workbench and to conduct a genome-wide association study of blood lipids on the pooled dataset aimed at demonstrating that biomedical researchers can be more productive when permitted to analyze the union of the datasets, as opposed to computing aggregate results in separate data silos for each dataset and then combining those aggregates. The project includes notebooks needed to perform a regenie genome-wide association study of lipids over the exonic variants within the All of Us alpha3 release of genomic data in a siloed fashion.
The “Replication of dissecting racial bias paper” demonstration project replicates the study, “Dissecting racial bias in an algorithm used to manage the health of populations” for an AIM-AHEAD workshop. The project aims to answer if we can predict the health status of participants in the next year using their health status and demographics and does the prediction for the health status differ by participant race and education level. The project includes notebooks on how to train a machine learning model that will predict the health status of a participant the year following the participant’s enrollment.
The “Geographic variation in obesity” demonstration project examines the quality and utility of the All of Us Researcher Workbench for accelerating precision medicine by replicating methods from existing studies that examine the prevalence of obesity at the population level. The project evaluates the measurements of obesity in the physical measurements data and the electronic health record (EHR) data using methods similar to the Ward et al. publication that assessed prevalence of obesity in the U.S. by state using Behavioral Risk Factor Surveillance System Survey data.
The “Genome-wide association study (GWAS) on LDL cholesterol with Regenie and dsub (version 6)” demonstration project recreated an efficient and scalable GWAS across whole genome sequenced data on an LDL Cholesterol phenotype in the All of Us Researcher Workbench.
The “Genome-wide association study (GWAS) on LDL cholesterol with Regenie (version 7)” demonstration project recreated an efficient and scalable GWAS across whole genome sequenced data on an LDL Cholesterol phenotype in the All of Us Researcher Workbench.
The “Genetic ancestry” demonstration project explores how genetic ancestry can affect health outcomes via differences in the frequencies of variants associated with disease and drug response.
The “Regenie LDL genome-wide association study (GWAS) using Cromwell” demonstration project uses Cromwell to run regenie via Workflow Description Language (WDL) for a notebook to analyze the regenie GWAS results. The project uses LDL cholesterol as the phenotype of interest and uses participant age and sex assigned at birth as covariates along with the top 15 ancestry principal components (PCs).
The “Social determinants of health” demonstration project reviews descriptive data analysis of the survey responses, psychometric analysis of social determinants of health scales, and item non-response rates across demographic variables.
The “Polygenic risk score genetic ancestry calibration” demonstration project aims to improve the ability to address the genetic ancestry-dependent bias in polygenic risk score (PRS) for 10 conditions: Asthma, atrial fibrillation, breast cancer, chronic kidney disease, coronary heart disease, hypercholesterolemia, obesity, prostate cancer, type 1 diabetes, and type 2 diabetes.
The “General health and wellbeing of sexual and gender minority (SGM) participants” demonstration project demonstrates the diversity and utility of All of Us Research Program by characterizing the demographics, health conditions, and behaviors of SGM participants.
The “Diabetes mellitus Medication Prescription Patterns” demonstration project explores a study designed to understand differences in prescriptions of newer generation diabetes medications such as GLP-1 agonists and SGLT-2 inhibitors by looking at patterns within the All of Us dataset. The project relates to the SAS analytics guide: SAS Analytics Guide - How to perform logistic regression.
The “Sociodemographic differences in treatment of mental disorders” demonstration project provides an example of statistical analysis processes used to assess concordance between self-reported lifetime depression diagnosis and depressive disorder diagnoses documented in available electronic health records (EHR) using survey and EHR data from the All of Us dataset. The project relates to the SAS analytics guide: SAS Analytics Guide - How to perform binary logistic regression.
The “Pharmacogenomics (PGx) variant frequency and medication exposures” demonstration project assesses the All of Us short-read whole genome sequencing (srWGS) variant data for the presence of specific alleles and predicted phenotypes known to be associated with adverse drug reactions or altered dosage recommendations. The project reviews frequencies of pharmacogenomics variants/haplotypes in All of Us Research Program participants. The specific Controlled Tier Curated Data Repository (CDR) paths related to this project can be found under srWGS PGx Haplotype Calls in the Controlled CDR Directory.
Phenotype library
Phenotype library workspaces demonstrate how computable electronic phenotypes can be implemented within the All of Us dataset using examples of previously published phenotype algorithms.
The “Breast cancer (Registered Tier)” phenotype library provides researchers with published phenotype algorithms for breast cancer. The library provides a supplement to the validated breast cancer phenotype by demonstrating a method to identify additional participants according to their answers to breast cancer-related questions on participant provided information (PPI) surveys and describes the query performed to capture a cohort according to a predefined algorithm for breast cancer. Phenotypes retrieved from the Phenotype Knowledge Base (PheKB).
Breast cancer (Controlled Tier)
The “Breast cancer (Controlled Tier)” phenotype library provides researchers with published phenotype algorithms for breast cancer. The library provides a supplement to the validated breast cancer phenotype by demonstrating a method to identify additional participants according to their answers to breast cancer-related questions on participant provided information (PPI) surveys and describes the query performed to capture a cohort according to a predefined algorithm for breast cancer. Phenotypes retrieved from the Phenotype Knowledge Base (PheKB).
Dementia
The “Dementia” phenotype library provides researchers with published phenotype algorithms for dementia. The library describes the query performed to capture a cohort according to a predefined phenotype algorithm for dementia. Phenotypes retrieved from the Phenotype Knowledge Base (PheKB).
Depression
The “Depression” phenotype library provides researchers with published phenotype algorithms for depression. The library describes the query performed to capture three cohorts according to a predefined phenotype algorithm for depression. Phenotypes retrieved from the Phenotype Knowledge Base (PheKB).
Ischemic heart disease (Registered Tier)
The “Ischemic heart disease (Registered Tier)” phenotype library provides researchers with published phenotype algorithms for ischemic heart disease. The library describes the query performed to capture a cohort according to a predefined phenotype algorithm for ischemic heart disease, also known as coronary artery disease. Phenotypes retrieved from the Phenotype Knowledge Base (PheKB).
The “Ischemic heart disease (Controlled Tier)” phenotype library provides researchers with published phenotype algorithms for ischemic heart disease. The library describes the query performed to capture a cohort according to a predefined phenotype algorithm for ischemic heart disease, also known as coronary artery disease. Phenotypes retrieved from the Phenotype Knowledge Base (PheKB).
The “Diabetes (Registered Tier)” phenotype library provides researchers with published phenotype algorithms for diabetes. The library describes the query performed to capture a cohort according to a predefined phenotype algorithm for type 2 diabetes according to four different cases. Phenotypes retrieved from the Phenotype Knowledge Base (PheKB).
The “Diabetes (Controlled Tier)” phenotype library provides researchers with published phenotype algorithms for diabetes. The library describes the query performed to capture a cohort according to a predefined phenotype algorithm for type 2 diabetes according to four different cases. Phenotypes retrieved from the Phenotype Knowledge Base (PheKB).
Community workspaces
Community workspaces foster knowledge-sharing, collaboration, and learning by allowing registered All of Us Researcher Workbench users to share their workspace with all other registered Researcher Workbench users.
For a full list of community workspaces, log in to the Researcher Workbench. To publish your workspace as a community workspace, read “Publishing a community workspace.”
Comments
0 comments
Article is closed for comments.