Researcher
Resources

The PMAP Cookbook is a textbook of computational notebooks to help investigators at Johns Hopkins conduct clinical research.

The Cookbook provides examples of how to work with clinical data within PMAP using a combination of R and Python programming languages and modern data science libraries.

Data from each PMCOE is ingested into PMAP and stored in the same location with all other Johns Hopkins data. PMAP is secure and HIPAA-compliant. Access controls (based on JHED IDs) limit clinicians and researchers to the data defined by their IRB approval.

You’ll need to fill the skill set of two key roles:

• Data management know-how to maintain data integrity and serve selected data to researchers.
• A biostatistician or data scientist to conduct data analysis.

Options include finding expertise in your team or department, or working with partners in CCDA, School of Public Health, APL, or others.

inHealth is continuously ingesting new data sources into PMAP. Review the Data Catalog (in the SAFE Desktop environment). If your data source is not listed, there is a process in place to present your case for inclusion. You can work with your PMCOE project manager or someone on the Precision Medicine team to submit a request for consideration. We formally evaluate, rank, and weigh new data sources according to the following criteria:

• Strategic Priority
• Research Efficiency
• Data Quality
• Generalizability
• Work Effort Required

Yes, find this glossary of common terms here.

The PMAP Data Catalog contains information about the data that are available in the PMAP. The research community can browse and search the contents of the Data Catalog that includes sources such as Epic, REDCap, open specimen data set, imaging data set, etc.

The Data Catalog does not show the actual patient data. Rather, it provides information about the available data to guide subsequent requests for that data. It is THE place to look for the most up-to-date information on PMAP data.

• For further information about using the Data Catalog, refer to the video tutorial here.
• For access to the Data Catalog (currently available in the SAFE Desktop environment)

Don't be intimidated when it comes to data science tooling. A great way to grow your data science skills is to learn from others. inHealth is expanding its team and resources in order to support researchers.

The Precision Medicine website was set up as a portal for all your clinical research needs in PMAP.

• The PMAP Cookbooks provide step-by-step details on big data computing within crunchr using Jupyter Notebooks.
• Events such as Crunch Time and the annual Johns Hopkins Precision Medicine Symposium serve as networking opportunities.
• Courses on Precision Medicine offered by Paul Nagy (search offerings through the Department of Health Sciences Informatics) provide hands-on learning opportunities.

A pilot is underway to establish an institution-wide eConsent process. If successful, this process will be implemented and available to all clinical research protocols approved by JH IRB (anticipated Q1 2021).

We determine this on a case-by-case basis. Please find the different levels of PM access at the conclusion of your PMCOE tenure and/or contact us to learn more.

Before beginning the IRB Approval Process, the following steps must be completed:

• Meet with inHealth leadership
• Determine human and financial resources
• Develop strategy and fix initial priorities
• Meet with the Institute for Clinical and Translational Research (ICTR) team designees
• Identify research resources (who will have access to your PMAP registry data?)
• Explore the PMAP Data Catalog
• Draft registry data specifications

You can find the new IRB application flow here.

inHealth is a strategic initiative to advance precision medicine at Johns Hopkins. By leveraging university-wide assets from the disciplines of medicine, engineering, nursing, public health, and our Applied Physics Laboratory, our team is developing patient-level insights that will improve health care.

The inHealth program is a part of the Johns Hopkins Medicine Strategic Plan to use data and technology in innovative ways to guide decision making in every corner of the Johns Hopkins organization.

The Precision Medicine Analytics Platform (PMAP) is a system that lets you obtain comprehensive clinical research data sets. The platform ingests disparate clinical and research data sources and provides machine learning tools to draw insights. In PMAP, the data collected for your studies is protected.

In the analysis layer of precision medicine at Johns Hopkins, several different environments are available:

• SAFE Desktop: A secure virtual desktop that emulates a traditional desktop experience while providing access to various desktop analytical tools such as SAS, Stata, RStudio, Anaconda, PowerBI, and NVivo. Learn more about SAFE Desktop.
• crunchr: A collaborative data science platform based on the leading data science platform, Jupyter. This environment comes packed with support for Python and R within JupyterLab and RStudio. For security purposes, Crunchr platform (crunchr.pm.edu) can only be reached from the Safe Desktop.

To request a SAFE desktop for your study team or to request access to an existing SAFE folder, please submit the request using the SAFE Desktop Request Form.

Go to Johns Hopkins Precision Medicine Analytics Platform (PMAP) Cookbook for detailed instructions on how to get to the crunchr website.

• Phoenix: A Linux-based High Performance Compute Cluster (HPCC) that can fit your most demanding analytical and processing needs.

There are several learning resources available while getting started on PMAP.

The Johns Hopkins PMAP Cookbook has links to many important FAQs and How To’s including:

• Before you start cooking
• Intro to crunchr and Jupyter Notebooks
• Accessing your data
• Working with EMR data
• Conducting Natural Language Processing (NLP) projects
• Working with medical images
• Programming resources

Access the various analytic environments here:

• Python: https://pm.jh.edu/cookbook/Py_01.html
• R: https://pm.jh.edu/cookbook/R_01.html
• Stata: https://pm.jh.edu/cookbook/STATA_01.html

Access the Data Catalog User Guide here OR view the Data Catalog video tutorial.

Access crunchr Support through SAFE Desktop here.

With the Precision Medicine Analytics Platform (PMAP), you can combine data from multiple sources, including REDCap. Researchers can bring in data from a REDCap database, and enrich it with data from Epic, or from other systems such as imaging or genomics.

When projects use SAFE Desktop or PMAP as intended, they’re considered Tier 1 Risk (lowest risk) by the Johns Hopkins Data Trust. This low-risk tier waives the need for a review by the Data Trust, unless data is being shared outside the institution. For this reason, most researchers don’t need separate approval from the Data Trust when they get IRB approval, skipping a step that sometimes takes months.

If you would like access to an existing research project in PMAP, you MUST request access through the PI.*. The PI needs to fill out this form and submit for access request: https://t.jh.edu/pmap-edit-team.

*The PI has a legal requirement to control access to study data for his/her research team.

If you are a researcher or a member of a study team and would like additional information, please visit the Support Resources on the PM Portal.

A de-identified dataset of asthma and diabetes patient records are available for PMAP training.

Submit this request form for access to the Training Dataset.