This repository is under review for potential modification in compliance with Administration directives.
The AMP PD Knowledge Portal was developed to host and share resources related to Parkinson’s disease research and remains fully operational. We continue to maintain and accept Parkinson’s disease and related disorders data and resources throughout this review process.

Approved Data Access Requests

First Name Last Name Institution/Company Proposed Analysis Updates to Research Use Statement Date Approved
Michael Greicius Stanford University We intend to use AMP-PD whole-genome sequencing and clinical phenotype data alongside external biobanks (UK Biobank, All of Us, and ADSP) to search for rare variants that affect risk across Parkinson's disease and related disorders. By applying quality control, functional annotation, and statistical association models (adjusting for age, sex, and principal components), our analysis plan aims to uncover novel risk genes and clarify missing heritability across neurodegenerative disease spectrums.

We intend to use AMP-PD whole-genome sequencing and clinical phenotype data alongside external biobanks (UK Biobank, All of Us, and ADSP) to search for rare variants that affect risk across Parkinson's disease and related disorders. By applying quality control, functional annotation, and statistical association models (adjusting for age, sex, and principal components), our analysis plan aims to uncover novel risk genes and clarify missing heritability across neurodegenerative disease spectrums.

Thomas Gasser University of Tübingen, Hertie Institute for Clinical Brain Research I am a founding member of the IPDGC consortium and currently on the steering committee of the GP2 project in the ASAP initiative. At the University of Tübingen we have genetic, biomarker and longitudinal clinical data on a local cohort of more than 1000 sporadic PD patients and 250 patients with rare genetic mutations (mostly GBA. We are analyzing our data presently to determine predictors of progression and would like to use AMP-PD data for replication. We also plan to use AMP-PD to contribute our cohorts to the GP2 consortium, depending on solving legal and regulatory issues of data protection.
Raquel Real University College London Parkinson’s disease (PD) is often complicated with cognitive impairment (MCI) and dementia, leading to an important loss of quality of life amongst individuals living with PD. Age is the main risk factor for the development of cognitive loss in PD, but several genetic risk factors have been identified that also contribute to the risk of developing dementia in PD (PDD). Among the most well-established genetic risk factors are the APOE4 allele and variants in the GBA1 gene. Two recent time-to-event GWAS studies of ~4000 individuals have identified additional genetic risk loci for the progression to dementia (Liu et al. 2021; Real et al. 2023), which have not yet been replicated. We aim to expand on these previous studies to investigate the genetic contribution for the progression to MCI and dementia in PD, using data generated within GP2 (https://gp2.org) from multiple longitudinal cohorts and deposited in AMP-PD. This is a GP2-led study that will be the largest of its kind to date. We aim to analyse data from ~10,000 individuals, thus significantly increasing the power to detect genetic associations to better understand the genetic make-up that increases the risk of dementia in PD.

Identification of the genetic factors that contribute to the incidence and rate of cognitive decline is still incompletely understood, partially due to the limited availability of longitudinal cognitive data for large scale genetic studies. We are leveraging GP2 data to perform the largest European PD dementia genome wide association study (GWAS) to date. Using data from GP2 release 12 with longitudinal cognitive data and four additional independent cohorts, we will investigate time to dementia using genome-wide Cox regression, adjusted for sex, age at baseline, and the first six genetic principal components, followed by meta-analysis. Genetic effects on longitudinal MoCA change will be investigated using linear mixed-effects models in GP2 Release 12 and two independent cohorts, followed by meta-analysis. Genome-wide significant loci will be functionally annotated using FUMA, and we will perform fine-mapping of significant loci. We will calculate a PD dementia PRS and investigate if it predicts rate of MoCA decline using linear mixed-effects regression. We will also investigate potential overlap between genetic risk factors identified in PD case-control GWAS and this PD dementia GWAS.

J Andy Henrie Datatecnica Develop and test ability to safely combine GP2 & ASAP CRN datasets for meta-analysis. ASAP CRN Cloud development and support team.

N/A

Ryan Davis University of Sydney We had an ASAP funded project that aims to use a small number of PPMI cell lines from SNCA and LRRK2 patients (in concert with PRKN lines from our own biorespositories) to generate iPSC lines and then use CRISPR to edit out the causative mutations and edit in the other causative mutations.

We also have a GP2 project to diagnose 1000 monogenic PD cases, as well as MJFF-funded collection and analysis of 500 atypical PD cases through GP2. We aim to find new disease genes, new classes of pathogenic variants, determine a diagnostic yield, develop analytical capability, and validate findings using in vitro functional validation. We will develop integrated analytical pipelines and new approaches to analysing genome data that will initially be assessed on our cohort, with intention to apply for access to broader GP2 cohort data for application of our tools and capability to identify new diagnoses and determine the utility of our analytical capability developments.

We require the WGS data for the patients of interest to be used in the project in order to compare with WGS data to be generated during the course of the ASAP-funded project to ensure the intended target variants are removed or added by CRISPR and monitor the introduction of unintended variants between the different iterations (primary line to iPSC to first CRISPR edit to second CRISPR edit).

This analysis will form the foundation of the project to ensure the outcome measures in the cell lines are related to intended alterations of PD causing variants and not due to unintended variant introduction.

We have been attempting to download data from the PPMI/LONI interface with most of the WGS files stalling before finishing their download. Professor Andrew Singleton suggested we register with AMP-PD to better access the required data.

While our previous engagement was with the ASAP program, for which we are still in the process of drafting a manuscript, we have moved our focus to the GP2 program. Our research objectives are to determine the diagnostic yield of WGS in monogenic PD cases, identify new disease genes, develop capability in genome data analysis and return results to patients in a validated clinical genetics approach. We will also be increasing diagnosis by functional validation follow-up in my lab. We are recruiting 1000 monogenic patients and an additional 500 atypical PD patients. We are using the GP2 facilitated genome sequencing and considering all variant types. Our analysis involves the use of predictive tools, integrated analysis pipelines, and prioritised variant functional validation All data will remain de-identified and collaboration with cohort leads will facilitate appropriate return of results if allowed.

Alexandra Zirra Queen Mary University London I have started a PhD at Queen Mary University of London in 2023. My work is on the clinical phenotype of PD in different ethnical groups from East London where we have a high percentage of people from South Asia (40%) and Afro-Caribbean descent (15%).

I am interested in progression of PD and would like to branch out into transcriptomics. Both genomics and transcriptomics use similar pipelines and this experience would be very useful.

I have been involved in a project looking at transcriptomic determinants of PD dementia last year with Dr Teresa Perinan, QMUL. We have presented this work at the GP2 Annual Investigators Meeting. We aim to continue this work by exploring the BioFind dataset in more detail.

Research Objectives The objective is to investigate the contribution of Alzheimer's disease (AD) genetic risk to cognitive impairment and dementia in Parkinson's disease (PD), focusing on genetic factors beyond the APOE region. The primary analyses have been completed using AMP PD/GP2 data from 3,454 individuals with PD. Continued access is requested to support additional analyses, address potential peer-review comments, and strengthen the robustness and interpretation of the findings. Study Design This is a secondary analysis of existing AMP PD genetic and clinical data. No new participant recruitment or data collection is planned. The study uses genetic variants, AD polygenic risk scores, and cognitive phenotypes, including MoCA and MMSE scores, alongside relevant demographic and clinical characteristics. Analysis Plan The completed analyses examined AD-associated variants, mapped AD genes, and AD-PRS associations with cognitive outcomes. Additional analyses, if requested during peer review, may include sensitivity analyses using alternative cognitive definitions, adjustment for additional clinical covariates, further investigation of exploratory findings, and validation of genetic association and polygenic risk score results. Research Use Alignment with Data Use Limitations AMP PD data will be used exclusively for the approved research objectives and in accordance with all applicable data use restrictions. Data will remain within approved research environments, will not be used to identify individual participants, and will not be shared with unauthorised parties. Findings will be reported in aggregate form. Continued access is required primarily to facilitate manuscript revision, respond to reviewer requests, and ensure the reproducibility and scientific robustness of the completed research.

Bruno Santos-Lobato Federal University of Pará I am currently a trainee interested in applying to get Tier 1 and 2 data access from the AMP PD and GP2 initiatives. Data analyses will be conducted with the support and guidance of the ‘GP2 Trainee Network’ which is part of the Global Parkinson’s Genetics Program and funded by the Aligning Science Across Parkinson’s (ASAP) initiative.
My goal is to conduct research in the field of Parkinson’s disease genetics. I aim to study the genetic risk factor for the development of levodopa-induced dyskinesia. I am also interested in performing research using available transcriptomic data.
As a trainee, I hope to grow the field of Parkinson’s disease genetics and have a positive impact to the research in my country. This will be possible by providing data access to members of the GP2 initiative and beyond.

None

Mridul Mehta DaCapo Brainscience Inc DaCapo has been founded by leading researchers from Harvard, Brigham & Women's and Memorial Sloan Kettering. The focus is develop therapeutics for CNS diseases with initial focus on Parkinson's disease.
We are currently using different datasets to validate our genetic findings and prospective models for patient stratification as we progress our first therapeutic program for sporadic PD.

We plan to use the jointly-called WGS data from AMP-PD to validate genetic drivers of PD progression which have been identified by our platform. We are also using this data to understand differences in genetic mutations across different cohorts, especially as these are related to inclusion/exclusion criteria of the different cohorts. We curate the WGS data, then conduct all research in-house using scripts and methods that have been custom-built in-house for this purpose.

1. Our research objective is to validate genetic targets we have identified as driving progression in Parkinson’s Disease. These genes do not work in isolation, and having created in-house protein-protein interaction networks, we would like to use the AMP-PD dataset to corroborate our findings and deepen our understanding of the relationships between genetic, transcriptomic, proteomic and clinical data as it relates to the genetic mutations of interest.

2. The study design consists of using the data to study variant allele frequencies, clinical progression of patients with select variants, and confounding factors that might affect such outcomes. We wish to also consider the differences between cohorts in AMP-PD to understand how different inclusion/exclusion criteria affect these analyses.

3. The analysis plan follows the design detailed above. All analysis is done in-house using custom scripts and programs created for this task. We have also developed in-house algorithms to curate, clean and filter genetic and other data (including ancestry and familial analyses), which will be applied to the data before to ensure robustness of results.

4. We intend to follow the data use agreement in its entirety, and given the nature of our algorithms and analysis tools (which are all built in-house), we anticipate complete alignment with the agreement.

Mike Nalls Data Tecnica + NIH Helping with AMP-PD + GP2 + SysBio.

None.

Asaf Nohi Tel Aviv University Much of the genetic risk for Parkinson’s disease is still unexplained, and most PD risk variants lie in noncoding regions whose target genes and cell types are poorly understood. Rare-variant analyses have so far focused on coding regions. My preliminary gene-set burden analyses in AMP PD found FDR-significant pathways (e.g., regulation of gliogenesis), which shows this approach can uncover new PD biology. I plan to extend the analysis to rare noncoding regulatory variants in AMP PD whole-genome sequencing data, prioritized with a deep-learning model fine-tuned on single-cell data from human brain and immune cells, and to look for genetic modifiers that explain why many LRRK2 and GBA1 carriers never develop PD. AMP PD’s large whole-genome cohort of cases and controls is essential for this work, which could reveal new risk pathways, improve risk prediction and point to candidate therapeutic targets.

Research objectives My goal is to identify new genetic risk factors, risk pathways and modifiers of penetrance in Parkinson’s disease. The work extends from rare coding variants to rare noncoding regulatory variants, and it aims to explain why many carriers of pathogenic LRRK2 and GBA1 variants never develop PD. Study design This is a case–control study that uses AMP PD whole-genome sequencing data as the discovery cohort. Candidate associations will be replicated in independent cohorts, including All of Us. A nested analysis will compare affected and unaffected carriers of LRRK2 and GBA1 variants. Analysis plan Extend rare-variant (MAF < 1%) burden testing at the gene and GO gene-set levels from coding regions to noncoding regulatory regions (promoters and UTRs). Analyses will be adjusted for sex, age and ancestry PCs and corrected with FDR. Prioritize noncoding variants with Sei, a deep-learning regulatory model fine-tuned on single-cell data from human brain and immune cells, to identify the cell types and regulatory programs involved. Test for modifier variants and pathways that separate affected from unaffected LRRK2/GBA1 carriers. Alignment with data use limitations All analyses are research on PD and related disorders, carried out within the approved AMP PD environment by approved users. I will not attempt to re-identify participants or redistribute individual-level data, and I will publish only aggregate results, with AMP PD acknowledged.

Aurelie de Rus Jacquet Laval University Update to the analysis plan: The analysis plan remains the same as the original plan.

Original analysis plan: We would like to use the AMP PD data to better understand the role of the blood-brain barrier (BBB) in Parkinson's disease (PD). Our research has found that the BBB appears compromised in people with PD, however there is a significant gap in our understanding of the causes, consequences, and mechanisms underlying BBB dysfunction. The AMP PD omics data (genomics, proteomics plasma and CSF) will be used to identify changes related to neurovascular function in PD, and this data will complement our in vitro and in vivo mechanistic investigations. Understanding how and why BBB is breached could lead to the identification of new drug targets to prevent or slow disease progression.

Research objectives. The objective is to identify Parkinson’s disease (PD)-related dysfunction altering brain barriers, which could play a role in disease onset and/or progression. Study design. To address the objective, the study design is divided into two main strategies. (1) First, our laboratory produces primary data using human iPSC-based models of brain barriers, which serve as a reference to prioritize the investigation of specific markers in AMP-PDRD biofluid or sequencing databases. (2) We perform statistical analyses of AMP-PDRD datasets, in particular plasma and CSF proteomics, and autoantibody profiling. Analysis plan. Data are analyzed using a combination of statistical methods that enable sample stratification according to clinical variables and correction for confounding factors. Statistical analyses are performed in R using RStudio. Longitudinal analyses are conducted using generalized estimating equations (GEE), implemented with the geeglm() function from the geepack package. Additionally, Spearman's rank correlation analyses and linear regression models are performed to assess associations between marker proteins. Research use alignment with any data use limitations. Our research is consistent with the permitted uses of AMP-PDRD data. The research aligns with the data use agreement, and we confirm that we do not attempt to identify or contact participants, access to controlled data is restricted to researchers approved by AMP-PDRD, and the controlled data is not shared with non-approved researchers. All data is handled securely in compliance with applicable institutional and AMP-PDRD requirements.

Teresa Kleinz Institute of Neurogenetics I work as an M.D. at the Institute of Neurogenetics in Lübeck under the supervision of Christine Klein and in the GP2 monogenic team and am involved in the harmonization of AMP-PD, MJFF and MDSGene data. The merging project necessitates that we have access to the data and can therefore request further data from different centers in a second step.

I intend to use the dataset being part/contributing to the following groups: Family Analysis Working Group, The Data Merger, one prodromal project, and the PTPA research project: Systematic evaluation of PTPA variants in the large, ancestrally diverse GP2 dataset will determine whether biallelic PTPA variants recur in individuals with juvenile-onset parkinsonism and intellectual disability, or whether the currently reported families represent exceptionally rare findings. Building on our family with two affected siblings carrying biallelic PTPA variants, this project will refine the genetic evidence for PTPA as a PD gene, define the broader phenotypic and allelic spectrum, and prioritize individuals and variants for downstream functional studies of PP2A-related neurodegenerative pathways. Whole-genome sequencing data and linked phenotype information from GP2 and AMP-PD will be used to investigate rare variants in PTPA. Small-variant callsets will be screened for rare coding and splice-region variants, with a particular focus on biallelic genotypes consistent with recessive inheritance. Candidate variants will undergo standard quality control, functional annotation, and prioritization according to predicted molecular consequence, allele frequency in relevant ancestry groups, and in silico pathogenicity predictions. For prioritized PTPA variants, sequencing quality and allelic configuration will be reviewed where possible, including phasing to distinguish potential compound-heterozygous from cis configurations. The frequency of rare, predicted damaging PTPA variants will be compared between individuals with PD and available controls, including ancestry-stratified analyses where appropriate. Available clinical data from variant carriers will be examined for variables including age at onset, family history, cognitive impairment/intellectual disability, and additional clinical features. Finally, identified variants will be compared with the homozygous missense variant observed in our family and with previously reported PTPA alleles to assess whether recurrent genotypic and phenotypic patterns support PTPA as a recessive cause of early-onset parkinsonism with intellectual disability.

Michael Turchin Bristol Myers Squibb We are interested in using the AMP PD data for new target discovery and early target development. Primarily we will use genetics in combination with the different omics data from AMP PD to investigate the degree to which putative drug targets for PD are supported by causal human biology. This will include, but not limited to, association analyses, finemapping, conditional analyses, causal inference, and various multi-omic integrative methods.

N/A

Erika Lam UCL Project looking at comparing genomic data in the AMP-PD samples - including variation in genes associated with atypical parkinsonian disorders, Lewy Body disease such as MAPT, GRN etc with our own cohort data of atypical Parkinsonian disorders. Of relevance to identify genetic variants associated with specific phenotypes that overlap with atypical Parkinsonian disorders.

Research objective: characterisation of early onset parkinsonism Study design: analysis of rare variant carriers in UK EOPD cohorts in known PD genes and analysis of unexplained cases and families to discover potential new genes. Analysis plan: family analyses and replication with collaborators and across wider GP2 dataset In alignment with data use limitations as stipulated

Kirtana Veeraraghavan Verily Will be helping to support AMP PD on Verily Workbench

NA

Sophie Meyer Global Parkinson's Genetics Program Queen Mary University of London AMP PDRD will help advance my research objectives of understanding more about the prodromal phases of PD and supporting the global expansion of smell testing in GP2 through:

Study design
- Regularly query numbers on VWB inform study design in the UK-based study I work on, by allowing us to identify and inspect genetic variant carriers.
- Querying smell test data availability within GP2 has allowed me to identify and conduct outreach with cohorts who have hyposmic individuals

Analysis plans
- Following the Nairobi bioinformatics workshop I have been using the VWB platform to support analyses plans for comparing non-manifesting carriers in the GP2 dataset
- the first planned analysis in my PhD will be looking at a GWAS of olfaction as a binary trait. I will be using the AMP-PD dataset and VWB platform to extract, clean and analyse the necessary variables for this core GP2 project.

Research use alignment and data use limitations
The research I conduct is aligned with, and funded by, GP2. My research is focused on expanding smell testing across the globe, particularly in areas that are under-represented in research. I can confirm that my use of AMP-PD data is within the scope of participant consent and complies with AMP PD's data access, confidentiality and security requirements. Further, I will not attempt to identify participants or use the data for purposes outside of the approved research.

AMP PDRD will help advance my research objectives of understanding more about the prodromal phases of PD and supporting the global expansion of smell testing in GP2 through: Study design - Regularly query numbers on VWB inform study design in the UK-based study I work on, by allowing us to identify and inspect genetic variant carriers. - Querying smell test data availability within GP2 has allowed me to identify and conduct outreach with cohorts who have hyposmic individuals Analysis plans - Following the Nairobi bioinformatics workshop I have been using the VWB platform to support analyses plans for comparing non-manifesting carriers in the GP2 dataset - the first planned analysis in my PhD will be looking at a GWAS of olfaction as a binary trait. I will be using the AMP-PD dataset and VWB platform to extract, clean and analyse the necessary variables for this core GP2 project. Research use alignment and data use limitations The research I conduct is aligned with, and funded by, GP2. My research is focused on expanding smell testing across the globe, particularly in areas that are under-represented in research. I can confirm that my use of AMP-PD data is within the scope of participant consent and complies with AMP PD's data access, confidentiality and security requirements. Further, I will not attempt to identify participants or use the data for purposes outside of the approved research.

Heidi Isaksen Helse Fonna In collaboration with my co-supervisors, Jodi Maple Grødem and Alexey Shadrin, we are working towards use of GP2 data for my women health project concerning Dementia with Lewy Bodies (DLB). A sex-stratified GWAS for DLB have not yet been perfomed. A systematic review (https://doi.org/10.1038/s44400-026-00123-6) on sex differences in patients with DLB highlighted the need to investigate the disease further, with less bias. To perform a sex-stratified GWAS, we will need approximately 800 DLB cases of each sex.

Depending on how fast we can access the data. If we can use the data for our next manuscript (polygenic risk and hazard scores in Dementia with Lewy Bodies, and whether there is a sex difference). The manuscript where we need the data is for a sex-stratified GWAS with DLB data.

Feruza Sultamuratova Al-Farabi Kazakh National University Research objectives: This study aims to investigate clinical, imaging, and biomarker factors associated with Parkinson's disease progression using the AMP PD dataset. The findings will complement an ongoing PhD project evaluating motor and non-motor manifestations of Parkinson's disease and support a better understanding of disease heterogeneity and patient stratification for advanced therapies, including deep brain stimulation (DBS).
Study design: A retrospective observational study will be conducted using de-identified longitudinal data available within the AMP PD platform. Demographic, clinical, neuroimaging, biospecimen, and genetic variables will be analyzed where available.
Analysis plan: Descriptive statistics will summarize baseline characteristics. Associations between disease severity, motor and non-motor symptoms, cognitive function, and available biomarkers will be evaluated using appropriate univariate and multivariable statistical methods. Longitudinal analyses will assess factors associated with disease progression over time. Findings will be compared with data from our institutional cohort to identify similarities and differences in clinical trajectories.
Research use alignment with data use limitations: The AMP PD data will be used exclusively for non-commercial academic research. All analyses will comply with AMP PD Data Use Agreements and institutional ethical requirements. Only de-identified participant data will be accessed, no attempts will be made to re-identify individuals, and results will be reported only in aggregate form.

-

Kostantin Kiianitsa University of Washington Research objectives
The objective is to investigate whether genetic variants in key neurodegenerative disease genes, particularly variants identified or implicated in Alzheimer's disease (AD) datasets, show different distributions in Parkinson's disease (PD). We are particularly interested in rare and low-frequency variants for which large cohorts are required to obtain informative allele counts.
Study design and analysis plan
We will use AMP PD/GP2 WGS Release 12 summary results (>63,000 individuals) to examine selected variants across PD, control, and “other” phenotype groups. We will obtain available allele count (AC), allele number (AN), and allele frequency (AF) and compare variant frequencies between groups. Where available, ancestry-stratified summary data will be used to reduce potential effects of population structure. AMP PD frequencies may also be compared with population reference databases such as gnomAD and with observations from AD datasets.
Variants showing potential enrichment or depletion will be evaluated using appropriate statistical methods, with rare observations interpreted cautiously considering allele counts and confidence intervals. The results will be used to prioritize variants and genes for subsequent genetic and functional studies.
Research use alignment
This research is focused on understanding genetic factors shared between PD, AD, and related neurodegenerative disorders and is consistent with disease-related biomedical research use. Analyses will use cohort-level summary genetic data, and no attempt will be made to identify or contact individual participants.
Howard Urnovitz FBB Biomed Research objectives: In our own PAXgene whole-blood RNA-seq libraries (PD and healthy controls), we have catalogued about 30,000 unannotated transcript models. The research objective is to use AMP PD whole-blood RNA-seq to validate whether our panels are reproducibly transcribed at stable genomic positions in a much larger independent population.
Study design: We will use our RNA Systems Intelligence platform for secondary analysis of existing AMP PD whole-blood RNA-seq (PPMI, PDBP, BioFIND and related cohorts), comparing PD, prodromal/at-risk participants and healthy controls, using longitudinal visits where available. No new participants are enrolled and no new samples are collected.
Analysis plan: We will quantify RNA transcription in AMP PD whole-blood RNA-seq and apply marker panels fixed in advance on our own cohorts, without refitting. We will test panel scores against diagnosis, and within each participant across visits against Hoehn and Yahr stage, MDS-UPDRS and cognitive scores. Performance will be reported as discrimination and calibration with confidence intervals, with age, sex, cohort and sequencing batch as covariates..
Research use alignment with data use limitations: Analyses are restricted to research on Parkinson’s disease and related disorders and will be performed within the AMP PD workspace by the personnel listed on the Data Use Agreement. We will make no attempt to re-identify participants, will not redistribute individual-level data, and will report results in aggregate, consistent with the AMP PDRD Data Use Agreement and any cohort-specific data use limitations.

Research objectives: In our own PAXgene whole-blood RNA-seq libraries (PD and healthy controls), we have cataloged about 30,000 unannotated transcript models. The research objective is to use AMP PD whole-blood RNA-seq to validate whether our panels are reproducibly transcribed at stable genomic positions in a much larger independent population. Study design: We will use our RNA Systems Intelligence platform for secondary analysis of existing AMP PD whole-blood RNA-seq (PPMI, PDBP, BioFIND, and related cohorts), comparing PD, prodromal/at-risk participants, and healthy controls, using longitudinal visits where available. No new participants are enrolled, and no new samples are collected. Analysis plan: We will quantify RNA transcription in AMP PD whole-blood RNA-seq and apply marker panels fixed in advance on our own cohorts, without refitting. We will test panel scores against diagnosis, and within each participant across visits against Hoehn and Yahr stage, MDS-UPDRS, and cognitive scores. Performance will be reported as discrimination and calibration with confidence intervals, with age, sex, cohort, and sequencing batch as covariates.. Research use alignment with data use limitations: Analyses are restricted to research on Parkinson’s disease and related disorders and will be performed within the AMP PD workspace by the personnel listed on the Data Use Agreement. We will make no attempt to re-identify participants, will not redistribute individual-level data, and will report results in aggregate, consistent with the AMP PDRD Data Use Agreement and any cohort-specific data use limitations.

Sydney Stiles Indiana University School of Medicine Our long-term goal is studying the neuroprotective role of glycine receptors (GlyRs) in neurodegenerative disease. Our preliminary data demonstrate that glycine, and striatal GlyRs regulate dopamine release, and glycinergic function is decreased during Parkinson's disease (PD) progression. Therefore, we aim to investigate the molecular mechanisms linking glycinergic signaling, dopaminergic dysfunction and PD progression using the AMP PDRD Knowledge Platform.
Aim: Determine whether alterations in GlyRs subunits and glycine transporters are associated with PD diagnosis, disease severity, and progression.
Study design: Conduct an observational analysis integrating available transcriptomic, genomic, proteomic, and clinical datasets from healthy and PD patients.
Analysis plan: Evaluate gene expression, pathway enrichment, and genotype/phenotype associations analyses to identify biomarkers associated with glycinergic signaling dysfunction. Also, evaluate relationships between these molecular biomarkers, motor impairment, and disease progression. All this analysis will be adjusted by sex, demographic and clinical history.
Research use alignment: All analysis will follow the approved AMP PD/PDRD Data Use Agreement.
Thanh Long Nguyen Broad Institute Our primary objective in requesting this data is to support visualizations of harmonized data from multiple Accelerating Medicines Partnership (AMP) programs, including AMP-PD, through the SysBio FAIRplex project. These tools are specifically designed to help researchers explore and interpret harmonized biological datasets from multiple AMP programs in a more intuitive and interactive way. By integrating individual-level gene count data, we aim to provide new functionalities within FAIRplex that will enable users to visualize patterns, compare cohorts, and identify meaningful biological signals more efficiently. Ultimately, our intention is to lower the barriers to data exploration, empower researchers with more actionable insights, and foster greater collaboration and transparency within the research community. Access to these datasets is essential for ensuring that the visualizations and analytical features we develop are accurate, relevant, and responsive to the needs of scientists working with high-dimensional omics data.
Christopher Wilson Junipero Therapeutics Research objectives: Junipero Therapeutics develops oligonucleotide therapeutics that raise expression from the intact GBA1 allele in GBA1-associated Parkinson’s disease. Because this dosage-restorative approach depends on two parameters, feasibility hinges on the magnitude of the deficit to be corrected and the non-carrier ceiling above which overexpression becomes a liability. We seek AMP PDRD and GP2 data to bound both empirically and to prioritize which GBA1 genotypes a therapeutic must address.

Study design: Cross-sectional, genotype-stratified analysis of existing participants. No recruitment, contact, or specimen use. Strata: coding GBA1 carriers (p.N409S, p.L483P, p.E326K), the intronic branchpoint variant rs3115534-G, other and compound carriers, and non-carrier PD and control groups.

Analysis plan: (1) GBA1 variant spectrum and allele frequencies across GP2 ancestry groups using whole-genome sequencing and NeuroBooster Array data, with explicit GBA1/GBAP1 pseudogene disambiguation. (2) Where transcriptomic data permit, compare GBA1 transcript levels by genotype to bound the activation window required. (3) Characterize rs3115534-G carriers, in whom branchpoint disruption lowers glucocerebrosidase activity from an otherwise intact coding sequence (Álvarez Jerez et al. 2024) and which is therefore directly addressable by an expression-raising modality. (4) Relate genotype to age at onset and progression to inform the target product profile.

Data use alignment. All analyses are Parkinson’s-specific and consistent with disease-related research use. We will work only with de-identified data in the approved cloud workspace, will not attempt re-identification or linkage to external datasets, will not redistribute any record-level data, and will acknowledge AMP PDRD and GP2 in all resulting publications and presentations.
Lindsey Kaschmitter FBB Biomed Research objectives: In our own PAXgene whole-blood RNA-seq libraries (PD and healthy controls), we have cataloged about 30,000 unannotated transcript models. The research objective is to use AMP PD whole-blood RNA-seq to validate whether our panels are reproducibly transcribed at stable genomic positions in a much larger independent population.
Study design: We will use our RNA Systems Intelligence platform for secondary analysis of existing AMP PD whole-blood RNA-seq (PPMI, PDBP, BioFIND, and related cohorts), comparing PD, prodromal/at-risk participants, and healthy controls, using longitudinal visits where available. No new participants are enrolled, and no new samples are collected.
Analysis plan: We will quantify RNA transcription in AMP PD whole-blood RNA-seq and apply marker panels fixed in advance on our own cohorts, without refitting. We will test panel scores against diagnosis, and within each participant across visits against Hoehn and Yahr stage, MDS-UPDRS, and cognitive scores. Performance will be reported as discrimination and calibration with confidence intervals, with age, sex, cohort, and sequencing batch as covariates..
Research use alignment with data use limitations: Analyses are restricted to research on Parkinson’s disease and related disorders and will be performed within the AMP PD workspace by the personnel listed on the Data Use Agreement. We will make no attempt to re-identify participants, will not redistribute individual-level data, and will report results in aggregate, consistent with the AMP PDRD Data Use Agreement and any cohort-specific data use limitations.
AINURA SHURINA West Kazakhstan Marat Ospanov Medical University We plan to use genomic and available clinical and demographic data from AMP PD to conduct a comparative analysis of genetic variants associated with Parkinson’s disease. AMP PD data will serve as an independent reference dataset for comparison with findings obtained in the Kazakh population of Western Kazakhstan, validation of previously reported associations, and assessment of the reproducibility of identified genetic signals.
The study will be conducted in full compliance with all applicable data use limitations. The data will be used exclusively for scientific research related to Parkinson’s disease, without any attempt to identify or re-identify individual participants.

We plan to use genomic and available clinical and demographic data from AMP PD to conduct a comparative analysis of genetic variants associated with Parkinson’s disease. AMP PD data will serve as an independent reference dataset for comparison with findings obtained in the Kazakh population of Western Kazakhstan, validation of previously reported associations, and assessment of the reproducibility of identified genetic signals. The study will be conducted in full compliance with all applicable data use limitations. The data will be used exclusively for scientific research related to Parkinson’s disease, without any attempt to identify or re-identify individual participants.

Hansruedi Mathys University of Pittsburgh We will use AMP PD/PDRD postmortem brain single-nucleus RNA-sequencing data and associated clinical, neuropathological and, where available and authorized, genomic information to identify cell-type- and brain-region-specific molecular changes in Parkinson's disease (PD). Our objectives are to characterize neuronal and glial alterations associated with PD and neuropathological severity, identify candidate pathways of cellular vulnerability and resilience, and assess which changes are shared with aging and Alzheimer's disease.

This is an observational secondary analysis of de-identified data from PD and unaffected donors. We will perform quality control, harmonize cell-type annotations, and analyze cell-type composition, donor-level pseudobulk differential expression, and gene-set/pathway enrichment. Statistical models will account for available demographic, technical and tissue-quality covariates and repeated sampling of donors across regions. We will control false discovery rates and assess reproducibility across regions and independent cohorts where feasible. Comparisons with separately authorized aging and Alzheimer's datasets will focus on gene- and pathway-level signatures. AMP PD/PDRD will provide an independent, deeply characterized PD resource to evaluate the reproducibility and disease specificity of findings from our human brain transcriptomic studies.

Use will be limited to approved research and applicable dataset-specific consent and data-use restrictions. Only approved users will access controlled data; we will not attempt re-identification or unauthorized redistribution. Data security, results sharing and publications will comply with the DUA and institutional requirements.
Julien Rey FBB Biomed Research objectives: In our own PAXgene whole-blood RNA-seq libraries (PD and healthy controls), we have cataloged about 30,000 unannotated transcript models. The research objective is to use AMP PD whole-blood RNA-seq to validate whether our panels are reproducibly transcribed at stable genomic positions in a much larger independent population.
Study design: We will use our RNA Systems Intelligence platform for secondary analysis of existing AMP PD whole-blood RNA-seq (PPMI, PDBP, BioFIND, and related cohorts), comparing PD, prodromal/at-risk participants, and healthy controls, using longitudinal visits where available. No new participants are enrolled, and no new samples are collected.
Analysis plan: We will quantify RNA transcription in AMP PD whole-blood RNA-seq and apply marker panels fixed in advance on our own cohorts, without refitting. We will test panel scores against diagnosis, and within each participant across visits against Hoehn and Yahr stage, MDS-UPDRS, and cognitive scores. Performance will be reported as discrimination and calibration with confidence intervals, with age, sex, cohort, and sequencing batch as covariates..
Research use alignment with data use limitations: Analyses are restricted to research on Parkinson’s disease and related disorders and will be performed within the AMP PD workspace by the personnel listed on the Data Use Agreement. We will make no attempt to re-identify participants, will not redistribute individual-level data, and will report results in aggregate, consistent with the AMP PDRD Data Use Agreement and any cohort-specific data use limitations.
Vaishnavi Sharma FBB Biomed Inc. Research objectives: In our own PAXgene whole-blood RNA-seq libraries (PD and healthy controls), we have cataloged about 30,000 unannotated transcript models. The research objective is to use AMP PD whole-blood RNA-seq to validate whether our panels are reproducibly transcribed at stable genomic positions in a much larger independent population.
Study design: We will use our RNA Systems Intelligence platform for secondary analysis of existing AMP PD whole-blood RNA-seq (PPMI, PDBP, BioFIND, and related cohorts), comparing PD, prodromal/at-risk participants, and healthy controls, using longitudinal visits where available. No new participants are enrolled, and no new samples are collected.
Analysis plan: We will quantify RNA transcription in AMP PD whole-blood RNA-seq and apply marker panels fixed in advance on our own cohorts, without refitting. We will test panel scores against diagnosis, and within each participant across visits against Hoehn and Yahr stage, MDS-UPDRS, and cognitive scores. Performance will be reported as discrimination and calibration with confidence intervals, with age, sex, cohort, and sequencing batch as covariates..
Research use alignment with data use limitations: Analyses are restricted to research on Parkinson’s disease and related disorders and will be performed within the AMP PD workspace by the personnel listed on the Data Use Agreement. We will make no attempt to re-identify participants, will not redistribute individual-level data, and will report results in aggregate, consistent with the AMP PDRD Data Use Agreement and any cohort-specific data use limitations.
Shih-Feng You Bristol Myers Squibb We plan to use snRNA-seq data from AMP PD to identify and validate differences in gene expression and gene regulatory networks between Parkinson’s disease (PD) patient brain tissue and control samples, and to examine how these differences relate to PD endophenotypes, such as motor subtype and cognitive decline. The study will employ a case-control design, comparing dopaminergic neurons and glia across PD, PDRD and control groups, with DEG and GRN as the main endpoint and adjustment for covariates. For the analysis, we will first perform quality control, normalization, and cell-type annotation, then infer gene regulatory networks using pySCENIC and compare network activity between groups and endophenotypes. Findings will be validated through independent cohort. All analyses will be limited to the approved research purpose.