Frequently Asked Questions

The Accelerating Medicines Partnership® (AMP®) program is a public-private partnership between the National Institutes of Health (NIH), multiple biopharmaceutical and life sciences companies, and non-profit organizations.

Managed through the Foundation for the NIH (FNIH), AMP unites resources of NIH and private partners to improve our understanding of disease pathways and transform current models for developing new treatments by:

• Identifying new targets, biomarkers and development paradigms
• Developing leading-edge tools and technologies
• Collecting large scale datasets and supporting analytics for open analysis by the research community 
• Generation of consensus platforms and procedures

Learn more here.

The AMP® PD program applies the successful public-private-partnership model of the Accelerating Medicines Partnership® (AMP®) initiative to bring together public partners that include the National Institute of Neurological Disorders and Stroke (NINDS), National Institute on Aging (NIA), the Food and Drug Administration (FDA)  with private-sector and non-profit partners that include the Aligning Science Across Parkinson's (ASAP) initiative, GSK, Pfizer, Sanofi, Bristol-Myers Squibb, Verily and the Michael J. Fox Foundation for Parkinson's Research (MJFF). This partnership works collaboratively to provide a deep molecular characterization and longitudinal clinical profiling of PD patient data and biosamples with the goal of jointly identifying and validating promising biomarkers and biological targets for therapeutics with the ultimate goal of increasing the number therapeutic options for those living with PD.  Learn more here.

A biomarker is a biological characteristic that can be objectively measured and evaluated as a signal of normal biological or pathological processes. 

Despite the significant advances made to understand the biological basis for PD, there continues to be limited success in disease modifying drug development. Through the AMP PD program design phase, government, non-profit and industry leaders identified the need to jointly identify and validate promising biomarkers for therapeutics with the ultimate goal of increasing the number therapeutic options for those living with PD

Before the start of the AMP PD program, the NIH and FDA with FNIH input helped produce BEST guidance to provide those working in this area to correctly identify biomarkers (analytes, anatomic features, or physiological characteristics) that have relevance to disease that can be reliably measured. BEST defines seven biomarker categories: susceptibility/risk, diagnostic, monitoring, prognostic, predictive, pharmacodynamic/response, and safety. While all biomarker categories are of importance, the AMP PD program is most interested in progression biomarkers. Learn more on each of the biomarker categories here. 

The AMP PD partnership is possible through the generous contributions from the AMP PD Unified Cohorts and the GP2 Federated Cohort. The AMP PD Unified Cohorts utilizes well characterized cohorts with existing biosamples and clinical data that were collected under comparable protocols. Through collaboration with these cohorts, new data are centrally generated from these cohorts’ biosamples in a standardized manner and mapped to harmonized common clinical data elements. All data are quality controlled in a uniform manner.  The GP2 Federated Cohort is a global genomic dataset that through collaboration integrates data from countries around the world.

Contributing study data benefits your study and the AMP PD program. 

When you contribute your clinical data, it will be harmonized with existing AMP PD data to facilitate search and discovery. This also provides your study greater exposure with links back to your study as the source of truth for detailed clinical data. Current AMP PD cohorts include clinical data and at least one set of DNA, RNA, or Proteomic assay data generated using a standardized approach. When you contribute this data, your study data will be integrated into the AMP PD sample catalog using a standardized nomenclature and will enjoy the benefits of a wealth of AMP PD program-developed analyses and versioned pipeline workflows.

You can find the criteria used by the AMP PD program here for determining whether data from a study is suitable for inclusion on the AMP PD Knowledge Platform. While we hope to accommodate your data sharing, there are no guarantees your cohort will be selected for inclusion.

If you want  to share information about available data for potential inclusion in the AMP PD program or have any questions, please contact admin@amp-pd.org. 

GP2 is a study for research purposes only. GP2 has provided a cohort submission form intended for Principal Investigators of research programs to share information about available data for potential inclusion in the GP2 study. GP2 is inviting form submissions from the start of the study and will follow-up to learn more about your proposed cohort.

If you have any questions, please contact GP2.

The AMP PD program does not manage a centralized biorepository of biosamples. Cohorts and their contributing studies manage their biosamples directly and independently.  Those interested in requesting biosamples are encouraged to follow instructions provided on each of the cohort sites. Some helpful links are provided below.

• The Parkinson's Disease Biospecimen Review Access Committee (PD BRAC) is responsible for reviewing all requests for PDBP, BioFIND, and LCC biospecimen access. 
• The PPMI Biospecimen Review Committee (BRC) is responsible for reviewing all requests for PPMI biospecimen access

The AMP PD program does not currently have a data coordinating center. Instead, AMP PD’s harmonized data decisions are derived through consensus with the contributing cohorts and AMP PD partners through our Committees, Working Groups and Sub-Groups. Technome works together with members to build out the AMP PD Knowledge Platform that leverages the technological infrastructure of Terra, a Verily and Broad Institute product, with their support.

GP2 federated data is independently generated, managed, and governed by GP2. 

The AMP PD program provides the research community access to AMP PD Harmonized Data and GP2 Federated Data through a single AMP PD Data Use Agreement. The AMP PD Harmonized Data consolidates longitudinal multi-omic and phenotypic information together from multiple cohorts into a single unified repository. The GP2 Federated Data brings together a one-of-a-kind global genomic dataset.

To see the latest data available from the AMP PD program and GP2 through the AMP PD Knowledge Platform under the single AMP PD Data Use Agreement, we encourage you to go to the Data Releases & Portal Updates on the AMP PD News & Update page. 

You may also find a summary of the data available on the summary data dashboards:

• AMP PD Harmonized Data Summary Data Dashboard
• GP2 Federated Summary Data Dashboard

To see more granular details on the AMP PD Harmonized Data refer to the following data dictionaries:

• Clinical
• Genomic
• Transcriptomic

To quickly identify the cohort, you can look at the participant identifier. The first two letters of the identifier indicate the cohort

• BF: BioFIND; 
• HB: HBS; 
• LB: LBD; 
• LC:LCC; 
• PD: PDBP; 
• PP: PPMI; 
• SY: STEADY-PD3; 
• SU: SURE-PD3

Baseline Visit: All AMP PD clinical data includes a baseline visit, which is defined as the participant’s first clinical evaluation as part of the cohort.*  All visit months are calculated relative to the participant's baseline visit (e.g., “Month 12” indicates data obtained 1 year after the baseline visit).   Other clinical data, such as diagnosis date, medication start date, etc. are calculated in months, relative to the baseline visit date, in order to allow researchers to establish a timeline for disease progression and treatment milestones.

For studies in which post-mortem samples were collected, such as the Postmortem Sequencing Data cohort, the date of death is used as a substitute for the baseline visit.  When premortem clinical data are available for a participant in a postmortem study, visit dates will be negative, indicating the number of months prior to death the event occurred.

*BioFIND participants had an initial screening visit, followed by a more comprehensive evaluation and biospecimen collection visit 2 weeks later.  Most BioFIND clinical and biospecimen data is therefore recorded as “Month 0.5”

There are currently two tiers of data access through the AMP PD Portal. Users in Tier 1 can access limited individual-level clinical data, omics summary results, and/or related metadata. Tier 2 allows users access to the full dataset, with individual-level clinical and omics data, as well as metadata.

Once users complete an online registration form and verify their email address, they will be asked to sign the AMP PD Data Use Agreement.  Users who are only interested in Tier 1 level data may electronically sign the AMP PD Data Use Agreement for access to Tier 1 level data. To access Tier 2 level data, the AMP PD Data Use Agreement must be signed by both the user and the user’s institutional official (on behalf of the institution under which they will conduct their research) in a .pdf format that can be submitted to the Access Compliance Team attached to the application or via a link within their email.

Users who want to access the limited AMP PD dataset (Tier 1) with the individual-level clinical data, omics summary results, and related metadata will need to complete the online registration form, sign the AMP PD Data Use Agreement, and verify their email address.

Users who want to access the full AMP PD dataset (Tier 2) with the individual-level clinical and omics data as well as metadata will need to complete the online registration form and must provide a signature from an institutional official who is authorized by that institution to legally bind the institution on all matters addressed in the AMP PD Data Use Agreement.   An example of such an official may be the designated institutional signing official for NIH eRA Commons users. 

• You can download, sign, and upload the signed version of the AMP PD Data Use Agreement when submitting your application. If you need to update your application information after you’ve submitted it, you can contact the AMP PD Access and Compliance Team (AMP PD ACT).
• Please note that failure to provide an institutional official signature will restrict access to all Tier 2 level AMP PD research data. By default, users will be assigned permissions to the "Limited Access" (clinical and summary data) tier until an institutional official signature on the AMP PD Data Use Agreement is provided. 

All AMP PD users were permitted access to GP2 summary aggregate data (Tier 1) beginning in December 2021. For AMP PD users to access GP2 individual-level clinical and genomics data as well as metadata (Tier 2), users will need to sign the updated AMP PD Data Use Agreement that includes the GP2 cohort requirements. If you are interested in providing an updated AMP PD Data Use Agreement, please contact the AMP PD Access and Compliance Team (AMP PD ACT).

You can update your information by contacting the AMP PD Access and Compliance Team (AMP PD ACT).

Prior to releasing de-identified data, the AMP PD program receives the following written assurances from the cohorts contributing data to the AMP PD Knowledge Platform regarding the shareability of their participant data: 

• Participant is fully consented for broad data sharing in a manner consistent with applicable laws enabling sharing with AMP PD users
• Participant is consented for broad individual level clinical and genomics data sharing with no restrictions outside what is stated in the Data Use Agreement. 
  • In some GP2 cases, individuals may not be consented for broad data sharing. When and if they are legally permitted to do so, GP2 will share only derived summary statistics abstracted from and not including participant level data
• Cohort Principal Investigator (PI) is informed on how data is being used and the PI is responsible for informing participants on this usage when necessitated by local law/regulations/or IRB to do so
• Data has no intellectual property limitations, country policy, legal, or other restrictions on sharing and use of participant data, including but not limited to
  • embargo period restrictions
  • data use restrictions on combining data from multiple sources
  • any restrictions that further restrict data use beyond the single AMP PD Data Use Agreement

No, the AMP PD Data Use Agreement was created with the input of all contributing cohorts and cohort funders.  It also has been reviewed at NIH to ensure compliance with all applicable federal laws, regulations, and NIH policies.  To ensure fairness and due to the number of stakeholders, no edits will be accepted from users. 

Many of the AMP PD data elements are available for download. Users who want to download these permitted data elements will be required to document why they are downloading the data to the Access Control Team, ACT@amp-pd.org and ACT@gp2.org. 

Yes, all users who download permitted data elements must agree to the policies outlined in the AMP PD Data Use Agreement. For the omics data it is up to the users and responsible institution to provide summary information about their analyses and maintain an updated Data Use Agreement on an annual basis. This is required to comply with the NIH Genomics Data Sharing Policy. The user and their institutional official should be aware that all obligations with respect to downloaded data that are described in the AMP PD Data Use Agreement (e.g., terms addressing non-disclosure, data security, etc.) survive expiration or termination of the AMP PD Data Use Agreement and remain in effect as long as the downloaded data is in existence.  

Of note, users who download data will also be responsible for the download costs. 

In order to create the best platform possible for sharing and collaborating in research, including best practices for data security, the AMP PD program encourages cloud analysis and seeks to understand use cases for data downloads.

As discussed in the AMP PD Governance and Policies and required by the AMP PD Data Use Agreement, users who want to download AMP PD omics data to local servers must ensure that the data cannot be accessed by anyone who does not have an up-to-date access approval from the Access Control Teams (ACTs).   Before downloading the data, it is a good idea to contact your host institution to verify that they are willing to accept responsibility for the security of the downloaded data, even after the AMP PD Data Use Agreement expires or is terminated.

Please let admin@amp-pd.org also know more about what types of analyses you intend to do; how you intend to work with the data. We would like to be sure that downloading the data is the right choice for your needs, as well as help ensure you download the data you need. 

Once the AMP PD team has confirmed that downloading the data is the right path for you, please reach out to admin@amp-pd.org for further instruction on using appropriate gcloud commands.v

To address the question about the size and egress cost of data, the AMP PD team would need to know specifically what data you're looking for.  Within the AMP PD harmonized Cohorts, you can find  research-ready WGS, RNASeq and clinical datasets available for your evaluation. The egress costs depend on where the data is being sent, but are generally between 10 and 25 cents per GiB.  Using the latest AMP PD Release, Release 2.5 dataset as an example, here is a summary of the size of each dataset, and the approximate egress costs:

• Clinical and Summary Metadata Release Datasets
  • Release 1.0: 24 MiB (negligible egress cost)
  • Release 2.0: 29 MiB (negligible egress cost)
  • Release 2.5: 31 MiB (negligible egress cost)
• RNASeq Release Datasets
  • Release 1.0: 450 GiB (approx. $45 to $115 egress)
  • Release 2.0: 456 GiB (approx. $45 to $115 egress)
  • Release 2.5: 465 GiB (approx. $45 to $115 egress)
• WGS Release Datasets
  • Release 1.0: 1.5 TiB (approx. $150 to $375 egress)
  • Release 2.0: 7.5 TiB (approx. $750 to $1,875 egress)
  • Release 2.5: 8.1 TiB (approx. $810 to $2,000 egress)
  • Release 2.5 WGS variant data is also available in a smaller Plink format: 640 GiB (approx. $64 to $160 egress)
• WGS Single-Sample Data
  • 10,432 subjects available with VCF, CRAM and Metrics data
  • 22.5 GiB average per subject
  • Approx. $2 to $5 per subject ($20,000 to $50,000 egress for all subjects)
• RNASeq Single-Sample Data
  • 8,461 samples available with Picard Metrics, Salmon Quantification, Star Align-Reads and Subread Feature-Counts data
  • 18 GiB average per subject
  • Approx. $2 to $4.50 per subject ($17,000 to $38,000 egress for all subjects)

For the AMP PD Unified Cohorts, the AMP PD Knowledge Platform hosts the harmonized omics data and clinical data elements only. Many cohorts contributing to the AMP PD program have funded data collection and generation outside of the AMP PD program framework. These additional data are accessible through the cohorts preferred data distribution platforms.

Since the AMP PD program shares back the cohort specific data it generates to each cohort, you may also access and analyze this generated data within the context of the cohorts. Acknowledgement of the AMP PD program is still required whether you access the data through the AMP PD Knowledge Portal or through the cohort’s platform.

We encourage users to explore cohort specific databases. 

For the harmonized data, the AMP PD team identified individuals using genetics that are likely to have participated in more than one cohort study. These were identified through quality control methods outlined here, with documentation mapping participant’s IDs across studies being made available within the AMP PD Metadata Tables ‘amp_pd_participant_wgs_duplicates’.

For these participants, only the highest quality omics data are retained in the database. The clinical data is available across studies.

Where available, Global Unique Identifiers (GUIDs) were added to the AMP PD Knowledge Platform. The GUID enables data to be associated with a research participant without exposing or transferring the research participant's personally identifiable information (PII). GUIDs are random alphanumeric characters that are not generated directly from PII. This capability provides two primary benefits. It allows data about a research participant to be accumulated across projects over time, regardless of where and when that data was collected. The GUID, in addition to the genetic duplicates checks, can help users know if there are individuals included in AMP PD data that participate in multiple cohorts.  As more GUIDs become available, these will be added to the AMP PD Knowledge Platform.

The AMP PD program encourages users to analyze and publish papers based on the AMP PD data.  As set forth in the AMP PD Data Use Agreement, such publications must reference the AMP PD program and the Cohorts that contributed data to it in the Acknowledgments section of the relevant publications. As users publish, the AMP PD Portal will be updated to add publications that are made known to us on the AMP PD News & Updates page under Recent Articles. If you have published a paper that uses AMP PD data and do not see it listed, please email manuscript@amp-pd.org. 

It is hoped that new users will build upon the research findings generated by the research community. To better enable such a collaborative approach, the AMP PD program also welcomes sharing of methods by way of github and notebooks to ensure reproducibility of results. If interested in sharing such resources, please email admin@amp-pd.org.

The AMP PD Knowledge Platform brings users a seamless experience through a familiar collaborative cloud-based resource and computing environment to connect AMP PD’s and GP2’s massive databases in a federated manner. 

The AMP PD Knowledge Platform components include the:

• Portal:  A public website for interested members of the research community to register and find resources 
• Terra: A data science platform for users to access and explore data, develop analytical resources, run machine learning pipelines, and collaborate in ‘Getting Started’ Workspaces with Jupyter notebooks 
• Google Cloud: The data as files in Cloud Storage and tables in BigQuery available to users via Terra, VMs, or directly.

In the context used on this site, the term “federated” describes the inter-operation of the two distinct, formally disconnected, independently governed AMP PD and GP2 databases that have different internal structures.

Terra provides in-depth guides on their website here for every step in the research journey, from finding and managing data to pipelining with workflows to interactive analysis (Jupyter notebook, RStudio or Galaxy).

• Account and billing
• Managing Access
• Managing Cloud costs
• Workspaces
• Data
• Workflows
• Cloud Environments Analysis
• Working with Containers (Docker)
• Troubleshooting

All AMP PD data is stored on the Google Cloud Platform (GCP) and can be accessed by approved researchers using both AMP PD Researcher Workbench and GCP-native tools.

Data is available as files in Google Cloud Storage and as tables in Google BigQuery.

If you are interested in learning more about how to use the Google Cloud, training opportunities are available on the Google Website here.

If you are an NIH-funded researcher with Cloud access via the NIH STRIDES program, additional courses can be accessed through Coursera on the Carasoft website here.

Because Terra runs on the Google Cloud Platform, all Terra costs, including storage, compute, and data egress, are ultimately billed via Google Billing Accounts.

You will need to provide billing information, but  may be able to start with a free trial that includes $300 in compute and storage credits. Click here to link directly to Terra support for additional details regarding free credits.

In order to be compatible with multiple cloud environments, institutional payment systems, and security requirements, the Terra interface does not directly display the Google Cloud Console Billing interface. Instead, Terra will connect a Google Billing Account to an associated Terra Billing Project. When using Terra, it is the Terra Billing Projects that you will see in the interface, and to which Terra will charge your usage costs. In order to create or clone a new workspace in Terra, you must have access to at least one Terra Billing Project.

Currently, there are two mechanisms potentially available to reduce your cloud computing costs. These are:

• The 90-day, $300 Free Trial: New Google Cloud and Google Maps Platform users may be able to take advantage of a 90-day trial period that includes $300 in free Cloud Billing credits to explore and evaluate Google Cloud and Google Maps Platform products and services. Click here to link directly to Google Cloud to learn more.
• The STRIDES Initiative: NIH and NIH-funded institutions have access to discounts and credits on cloud services and tools. An agreement between your institute(s) and STRIDES is required.  Your institution may already have an agreement in place, so ask your leadership to find out. Learn more here.

A critical program goal of any AMP initiative is to advance research by creating a community resource of data, analyses, and other research resources which are publicly available and accessible to the broad biomedical community. To maximize scientific exchange and accelerate research in that field, it is expected that all information, data, protocols, resources, and methods developed by AMP investigators will be shared in a rapid and timely way with other investigators in the consortium and with the research community at-large. All data and tools generated will be deposited in a timely manner into the AMP PD Knowledge Platform or its federated platforms and made accessible for use by the broad biomedical community.

Depending on the data type (individual level or summary data, tissue type, and country of origin, etc.) data may be submitted to the AMP PD program and GP2.

The AMP PD and GP2 IP terms are explicitly listed in the AMP PD Data Use Agreement. These IP terms aim to ensure the data generated and shared remain unencumbered by IP, thus enabling the broadest possible vehicle for data-driven discovery. These policies also ensure that discoveries made on biological processes underpinning the data also remain unencumbered by IP.

Provision number sixteen (16) in the AMP PD Data Use Agreement states as follows:

I acknowledge that the AMP PD program and GP2 have separate intellectual property policies and agree as follows:

1. Data from the AMP PD program were generated under and are subject to an existing arrangement that has the following AMP PD Intellectual Property Policy that states: “AMP PD users agree not to file patent applications on research discoveries made using the AMP PD Data, except in the rare instance when a consensus of the Foundation for the National Institutes of Health (FNIH), the AMP PD Steering Committee and the AMP Executive Committee agree that it is in the best interests of the partnership and public health to do so. Intellectual property developed under National Institutes of Health (NIH) awards are subject to applicable Federal law, regulation, and NIH policy.”  Accordingly, it is in the rare instance that the AMP PD Steering Committee, through an approval protocol, will deem that it is in the best interest of the AMP PD program and the public health to grant an exception. If an exception is granted, I agree to grant the funding partners of the AMP PD program a nonexclusive, worldwide, royalty-free, sublicensable license to use and/or disclose the intellectual property rights in and to the research discoveries made using the AMP PD Data for noncommercial research purposes.
2. GP2 has been established as a pre-competitive consortium. All rights to pre-existing IP that is used in connection with the GP2 data will remain with the owner of such IP. GP2 will not acquire any rights in or to any pre-existing IP or improvements thereto. I agree not to file a patent application, copyright, or assert trade secrets or other IP on research discoveries (including improvements to pre-existing IP) made using data from GP2 (except that articles divulging research discoveries may be copyrighted). For example, if I create or improve an algorithm using GP2 data, I will not claim IP on this algorithm. (An “improvement” is a modification to, variation of, or derivation from the original IP.) For more information, please email act@gp2.org.
    

No, this would not be compatible with the AMP PD Data Use Agreement IP terms.  Additionally, in Association for Molecular Pathology (AMP) v. Myriad Genetics Inc., the United States Supreme Court ruled that human genes cannot be patented in the U.S. because DNA is a "product of nature." The Court decided that because nothing new is created when discovering a gene, there is no intellectual property to protect, so patents cannot be granted.

No, this would not be compatible with the AMP PD Data Use Agreement IP terms. In line with the above, our stance is that if nothing new is created when discovering a “biological target”, there is no intellectual property to protect.

Yes, any materials, methodologies, processes, computer software and code, standard operating procedures, or intellectual property owned and used by the Organization before AMP PD data use will remain that organization's property. Furthermore, the organization retains and shall own as its own property all future rights, title, and interest in and to all inventions, processes, technology, know-how, trade secrets, improvements, other intellectual properties, and other assets that have been developed by the organization during the course of AMP PD data use which does not incorporate or require the use of AMP PD data or the Results. 

Maybe, if analysis of AMP PD data pointed to something that you would like to develop separately from AMP PD and that development is not using AMP PD resources, then you can patent that product of your separate effort.  In addition, in very rare cases, when a discovery results directly from analysis of data from the AMP PD program, the AMP PD Steering Committee may grant an exception upon request and allow a patent application to be filed, if this is deemed to be in the best interest of the partnership and public health.

Maybe, if analysis of AMP PD data pointed to something that you would like to develop separately from AMP PD and that development is not using AMP PD resources, then you can patent that product of your separate effort. In addition, in very rare cases, when a discovery results directly from analysis of data from the AMP PD program, the AMP PD Steering Committee may grant an exception upon request and allow a patent application to be filed, if this is deemed to be in the best interest of the partnership and public health.

• Users that benefit from the data and tools in the AMP PD Knowledge Platform shall acknowledge resources accessed as per the AMP PD Data Use Agreement. Users agree to submit  manuscripts to the AMP PD Publications Committee (PC) at manuscript@amp-pd.org  two weeks prior to planned journal submission for review of AMP PD authorship citations.  Response from the AMP PD PC will be within 2 business days after receipt of the manuscript by the AMP PD PC members.  Non-response within the 2-business day window will indicate AMP PD PC’s concurrence with the authorship information provided in the manuscript. Authors are required to notify the AMP PD PC of acceptance or rejection of the manuscript and to provide the AMP PD PC with the manuscript citation and, if possible, URL to the article, upon acceptance.
• NIH and FNIH-funded investigators and institutions shall acknowledge funding from the FNIH and the Accelerated Medicines Partnership PD initiative in any paper(s) in which the data were published as a result of the award.

No, the AMP PD Data Use Agreement was created with the input of all contributing cohorts and cohort funders.  It also has been reviewed at NIH to ensure compliance with all applicable federal laws, regulations, and NIH policies.  To ensure fairness and due to the number of stakeholders, no edits will be accepted from users. 

Please contact the admin@amp-pd.org for further assistance. 

1. Feedback from Helpdesk Tickets
2. Working Group Members
3. Usability exercise