Introduction to Research Data Management for Life Sciences

Course goals

At the end of this course, students will be able to:

1. Design reproducible research workflows that address the social, ethical, and institutional challenges of data sharing, using the FAIR principles and tools like the Open Science Framework (OSF).
2. Construct machine-actionable research outputs by compiling metadata with controlled vocabularies, integrating file formats such as JSON, and publishing datasets via repositories such as Zenodo .
3. Develop modular and reproducible data analysis pipelines by creating and querying relational databases with SQL and integrating these workflows with Python scripting.

Content

Content
Is your perfectionism holding you back from sharing your data? Do you have files named AAAA_final_FINAL_v2.xls? Then Introduction to Research Data Management is for you. This course offers practical and conceptual insights into managing research data in the era of open science and AI. It begins with the common challenges researchers face: scattered datasets, inconsistent file naming, and uncertainty around sharing. Students explore how reproducibility and collaboration suffer under such practices and learn how to mitigate these issues using tools like OSF, FAIR data principles, and metadata standards. Throughout the course, students gain technical skills for managing databases, file formats, and building analytical pipelines that can stand the test of time far beyond individual research projects. These skills are applied to integrate, query, and publish structured data, ensuring that workflows are modular, transparent, and future-proof. The course is grounded in an engaging scenario-based format: a simulated disease outbreak in the fictional country of FAIRhaven, where students investigate, model, and publish data in collaboration with roles like data stewards and engineers.

Structure and Audience
This course is designed for lifelong learners, PhD candidates, and Master’s students in the life sciences, biomedical sciences, or biology. Two formats are available:

• A 4,5 EC version that includes a capstone project and is recommended for Master’s students or anyone with sufficient time.
• A 3 EC version that covers the core modules without the final project, suitable for PhD candidates or learners , only available for PhD candidates or learners with limited availability, however not for master students.

Both versions span 10 weeks, with 2 or 3 weeks of active online coursework, respectively. Participants are expected to have prior experience with programming (preferably in Python) and to be comfortable using the command line (terminal), as software installation and environment management will be done independently.

Course Modules
The course is structured into the following modules: Research Data Management principles (OSF, data lifecycle, reproducibility crisis); FAIR data and stakeholder roles; metadata, file formats (e.g. JSON), vocabularies and ontologies; CRUD operations and relational data modeling using SQL; modular analysis pipelines integrating SQL and Python (ETL); publishing data and metadata (ARGOS, Zenodo); and AI-readiness, ensuring machine-interoperable, reusable data. Throughout the course, we revisit the Harvard Medical School Bio-medical Data Lifecycle.
Assessment
Assessment consists of:

• Quizzes (20%), with three attempts per quiz and a minimum required score of 6
• Assignments (20%), with no minimum grade required, though resubmission is only permitted after consultation with the course coordinator
• A capstone assignment (60%), requiring a minimum grade of 5.0, with one opportunity for resubmission if the final grade is below 6.

Minimum final grade to pass the course
5.5

Literature and Study Material
All course materials are provided via Brightspace. Students are required to install and use software locally on their own devices. This includes: Anaconda, Python, Jupyter Notebook, and SQLite3.