CMSC 829C: Advanced Topics in Bioinformatics and Computational Biology

Lectures in the first few weeks

The instructor gives lectures introducing biological data analysis and the common algorithmic, machine learning, and hardware approaches used across applications, together with explicit instruction on how to read, review, analyze, and very clearly present a research paper. These sessions provide the fundamentals for the topics we cover for the rest of the term. Depending on enrollment, the number of lectures may be adjusted.

Paper presentations for the remainder of the term

Each session covers one research paper, led by a student. We release a list of papers spanning algorithms, machine learning, and hardware accelerators for bioinformatics, and you choose from this list based on your interests. With one paper per session, each student typically leads one session, and we try to schedule related papers in consecutive sessions.

• Prepare with a mentor. Before you present, you must meet at least once with your mentor, the instructor, at least one week in advance, bringing a draft of your slides. Acting on this feedback is part of your grade. Schedule this meeting early, because last-minute meetings leave no time for follow-up.
• Follow the talk structure. Cover the title, authors, and venue, then the problem and why it matters, the key ideas and insight, the mechanisms behind how it works, the key results, the strengths, the weaknesses, your own ideas on how it could be improved, and finally the takeaways and questions that open the discussion.
• Lead the discussion. In a 75-minute session, plan for about 40 minutes to present and analyze the paper and about 30 minutes to lead the discussion. Submit your final slides at least 24 hours before class.

Reading and discussion

For each presentation session you are not leading, read the paper in advance and submit discussion questions the night before.

Paper review at the end of the term

Late in the semester you write a structured paper review of a paper, following the way real conference and journal reviews are written. By then you will have spent the term reading and analyzing papers with a structured method, which is the method you will be using. We discuss the reviews together in the final session.

• Before-class preparedness10%You contacted your mentor at least one week before your session, brought a draft of your slides, and addressed the feedback you received.
• Presentation15%The clarity, structure, and content of your talk, how well you understand the paper, your delivery, and how well you answer questions. Understanding the paper covers how you frame the biological problem, the data and benchmarks, and the method's assumptions and limitations. Grading takes into account the difficulty of the paper, the time you had to prepare, and how well you addressed your mentor's feedback.
• Analysis and discussion leadership10%The depth and quality of your analysis (strengths, weaknesses, and your own ideas) and how well you lead the discussion.

We look at how actively and constructively you take part in the sessions you are not leading, meaning how you engage with the presenter and your peers, build on what others say, and help drive the discussion, often from the questions submitted beforehand.

Before each session you are not leading, read the assigned paper and submit thoughtful questions that can spark discussion. The presenter may share some of them during the discussion. Questions are graded on quality and on evidence of critical thinking, and generic questions such as "What if you ran it on dataset X?" are generally discouraged. To allow for busy weeks and the occasional absence, your five lowest discussion-question scores are dropped, including any sessions you could not attend.

You write a structured paper review of an assigned paper, following the way real conference and journal reviews are written, with a summary, the strengths, the weaknesses, concrete suggestions, and an overall assessment. We provide detailed instructions and the components to include. Your review is graded on content, clarity, and the quality of your analysis, and you get feedback to help you improve your reviewing skills. We discuss the reviews together in the last class.