The Biomedical Informatics and Cyberinfrastructure Enhancement Core offers the following services and support to Rhode Island investigators. These are done in coordination with the COBRE CBHD Computational Biology Core, Computing and Information Services, and Brown University Library:

URSA Initiative

A signature effort is the Unified Research data Sharing and Analysis (URSA) Initiative (formerly Unified Research data Sharing and Access). The overall goal of the URSA Initiative is to make data accessible and usable for research purposes by the Advance-CTR community through establishment of a shared technical infrastructure and common processes.

This initiative is coordinated by the Brown Center for Biomedical Informatics (BCBI) that operates the Advance-CTR Biomedical Informatics Core in close collaboration with information services, compliance programs, and research offices at Brown and Advance-CTR data partners: Lifespan, Care New England, Providence VA Medical Center, Rhode Island Quality Institute (RIQI), and Rhode Island Department of Health (RIDOH). For sensitive data, URSA utilizes the Stronghold server system at Brown (“URSA Stronghold”), which offers a secure computing environment for storing and analyzing data.

This effort consists of: 

  • URSA Cyberinfrastructure: Servers (Linux and Windows), databases, and network connections both within and outside the secure Stronghold computing environment at Brown as well as with health data partners (e.g., Lifespan and Rhode Island Quality Institute). In particular, URSA Stronghold is actively being used for analysis of electronic health data using general purpose programming languages like Julia and Python, as well as statistical programming languages such as R and SAS. The URSA Cyberinfrastructure continues to be enhanced and expanded to support the computing needs of investigators while ensuring privacy and security requirements designed to meet institutional, state, and federal requirements.

  • URSA Data Requests: Aggregate statistics and datasets (de-identified, limited, identified, or linked) for supporting the breadth of research studies. Local data sources include Electronic Health Record (EHR) systems (Lifespan, Care New England, and Providence VA Medical Center), CurrentCare Health Information Exchange (Rhode Island Quality Institute), and HealthFacts RI All-Payer Claims Database (Rhode Island Department of Health). In addition, efforts are being made to centralize and coordinate access to publicly-accessible datasets (e.g., from the National Institutes of Health, Food and Drug Administration, and Centers for Disease Control and Prevention). See more information below. 

  • URSA Governance: Responsible for formalizing and enforcing policies, procedures, and processes for using the URSA Cyberinfrastructure and fulfilling URSA Data Requests. This governance structure includes stakeholders from health data partners, COBRE Center for Computational Biology of Human Disease (CBHD), and Computing and Information Services (CIS) at Brown. 

  • URSA Computational Tools: Available or under active development for processing and analyzing data from diverse sources (see BCBI Software and Web Apps). This includes open-source or publicly-accessible tools for de-identification (DeIdentification.jl), machine learning (PredictMD.jl), and natural language processing (e.g., MetaMap).

Data Requests 

Facilitate requesting, accessing, and using data from health data partners and other data sources.

Local data sources include:

Publicly accessible data sources include (* local version available):

Data Storage and Management

Identify options and solutions for storing and managing data from health data partners or other data sources. This includes use of the URSA Stronghold computing environment for sensitive data, database management systems (e.g., MySQL), and established data standards.

Data Processing and Analysis

Use or develop computational tools for processing (including de-identifying) and analyzing data from diverse sources. This includes using open-source languages (e.g., R and Julia) for:

  • Data Mining, Machine Learning, and Deep Learning

  • Natural Language Processing

  • Visualization

Expertise with bioinformatics tools such as Ingenuity Pathway Analysis (IPA) is also provided.

Data Submission and Dissemination

Support archiving, sharing, and disseminating research products using best practices and available resources (e.g., Brown Digital Repository).

Software Requests 

Offer guidance in available software to support project needs and facilitate installation of software in research computing environments such as URSA Stronghold.

Software and Application Development

Provide programming support for custom development for creation and analysis of custom datasets or databases (e.g., using Julia and MySQL) or web applications (e.g., using React/Node.js or SMART on FHIR).

Grant Support

Offer guidance for projects requiring informatics and data science expertise, with a focus on support for developing extramural or pilot project applications.

Education

Semester-long and short courses are available including the 5x5 Clinical Informatics Short Course: 5 Topics and Computing Skills in 5 Days for medical residents/fellows and junior faculty. This course is offered periodically and provides hands-on experience with essential computing skills for electronic health data analysis using the URSA Cyberinfrastructure. The knowledge and skills obtained through this course will prepare participants for more advanced one-day sessions and other courses offered throughout the year (e.g., in statistics, machine learning, natural language processing, and visualization). 

See a full listing of courses along with descriptions and syllabi for more information. The next offering of the 5x5 Clinical Informatics Short Course is planned for Winter 2019/2020. Subscribe to the Advance-CTR email newsletter for updates and announcements.