Independent Study in Biomedical Data Science (ISBDS)

Introduction

This independent study course involves the design, execution and presentation of a research project, reflecting the focus on independent research appropriate for graduate study. The course may be taken for 1-3 credits according to the scope of the project. Students are expected to choose among the many relevant domain areas, find appropriate datasets, learn skills and background knowledge as needed, pose one or more research questions, then design and conduct a workflow and study to answer those questions, and finally, present these findings via report and virtual poster session.  This process is analogous to research efforts at all levels.  This course is independent in two distinct ways:

  1. The student must propose a research project in sufficient detail and receive instructor approval prior to registration.
  2. The student will execute the research project independently, with limited guidance from a designated advisor.

Milestones with approximate due dates:

  • Preliminary Research Plan: Pre-registration
  • Final Research Plan: Week 5
  • Progress report: Midterm
  • Poster: Week before finals
  • Final Report: Finals week

Grading policy

Research plan:10%
Progress report:10%
Poster:20%
Poster presentation:5%
Participation (online forum):5%
Final report:50%

Considering your background and interests

This course is designed for students with diverse backgrounds, skills and interests, so the projects can also be quite diverse. If you are a strong programmer, feel free to showcase that in your project. If you are stronger in statistics and data analysis, you may focus on that. If you have some particular domain knowledge or interest feel free to emphasize that. The course may be taken for between 1, 2, or 3 credits, to accommodate a range of project scope. While data science always involves analysis of data, there are many approaches and methods. However, understand that in this course we define data science to include quantitative, mathematical facts and analysis. Conceptual analysis will not suffice.

Finding research ideas and advisors

Numerous project ideas, templates, and resources are available spanning a variety of biomedical domains, from which students may choose, combine, and be inspired to generate new ideas. Examples of project reports and posters are available for guidance. The course instructors will endeavor to assign an appropriate advisor to each student, depending on the research topic and availability of advisors, however, limited advisors are available, and students are welcome and encouraged to find and solicit suitable, qualified advisors. 

Research project advisors

Advisors may be UNM faculty, staff, graduate students who have qualified for PhD candidacy. External advisors are permitted with instructor approval. The student and advisor should develop a clear understanding of roles and expectations at the beginning of the project, which should be consistent with the capabilities and availability of the advisor. As a rule of thumb, advisors are expected to commit one hour minimum per week per student. 

Templates for research plans

For this course we are employing templates to guide students in developing their research plan, and to facilitate the sharing of research ideas. We are producing and compiling a library of templates in a variety of areas and with varying specificity, as starting points for research plans.  Developing the preliminary and final research plan is designed to provide feedback to the student and assure the projects are reasonable in subject and scope.  The revision process mirrors the submission of abstracts for evaluation such as for conferences. As with scholarly articles and presentations, the title and abstract are very important in providing a summary to your audience. Additionally, the research plan must include the definition of the data sources and datasets to be analyzed, the research questions to be addressed, the computational tools and methods to be used, and any planned deliverables such as software components. Abstract length should be approximately 1 page.

Research plan approval process

Instructor approval for the preliminary research plan submitted by the student is a strict prerequisite for registration. This is to help ensure student success and projects at a high academic standard. Criteria for approval will include whether previous coursework is sufficient, whether the plan is scientifically valid and feasible with the available resources, and whether the scope is appropriate for the credits requested [1-3]. In accordance with commonly accepted practices in higher education, and federal regulations, each credit hour requires a minimum student effort of 3 hours per week for 15 weeks (45 hours). Revisions to the plan may be required prior to approval.

Virtual Poster Session

Poster presentations are a very common and effective way of communicating research findings to colleagues.  This assignment is intended to give you experience in this important mode of communication, by converting and condensing content from your final project into poster form, and recording a brief audio narration of your poster.  Typically it will be most efficient to generate your poster in parallel with your report, with selected figures and text.  The poster is an important product, more so your ability to explain your project orally while using the poster as a complimentary visual. Moreover, the ability to answer critical questions from an audience, whether in real-time at a face-to-face poster session or via online video and discussion.

Final Reports

The final report should reflect all the work you have done on the project, be fully self explanatory, and submitted as a single file (normally PDF).  (Supplementary materials are permitted.)  The content and style should resemble a scholarly article, incorporating text and appropriate figures to describe your methods and results, and provide sufficient context and interpretation relative to the biomedical domain.  In other words, your report should be comprehensible and meaningful to a biomedical scientist.  Here are some additional suggestions:

  1. Organize your work: all the data, code, notes, references, results.  If you are in a team, this may require meetings and collaboration tools (e.g. GDrive or Dropbox).
  2. You may wish to review some examples of scholarly articles, or  technical reports, for inspiration and guidance.
  3. Start early!  Begin writing early enough to allow time for problem solving and seeking help as needed.

Option: Working in Groups

You may choose to do a project in a group. Expectations as to effort for a group are correspondingly higher than for an individual. The report should clarify the contributions of each group member, in the style of a journal article. Regardless of your role in the group project, you must understand the project in full, and be able to answer questions about any aspect of the project. Each student is individually responsible for presenting the poster and answering questions regarding any portion of the project. 

Course support and discussion forum

The ISBDS Course Piazza online forum will be an essential and central resource for this course. Students, advisors and instructors will be encouraged to ask and answer questions via the forum, developing a valuable source of answers to frequent questions. Participation and contributions to the forum will be considered in your final grade.

Project ideas and templates

You are welcome to select from the project suggestions listed in the separate document Project Ideas and Templates, or do something else. The course GitHub repository is https://github.com/unmtransinfo/ISBDSCourse, for project templates, example code, documentation and other resources.  Students are encouraged to share additional project suggestions via the course online forum.

Community resources

This course is intended to provide a space for a community of biomedical data science students, instructors, and researchers. Through the sharing of research ideas, completed projects, and discussions, we hope to foster learning and collaboration. Past and present students, instructors, and advisors may choose to be included in public lists of participants, and may choose to publicly share projects and other resources, to foster outreach and engagement. 

Course leadership

  • Director: Tudor Oprea, MD, PhD, Professor and Chief of Translational Informatics Division
  • Instructor: Jessica Binder, PhD, Post-doctoral fellow, Translational Informatics Division
  • Instructor: Jeremy Yang, BEng, MA, Senior Research Scientist, Translational Informatics Division