392144 Privacy in Pangenomics

Seminar (Winter 2026/2027)
Genome Data Science group, Faculty of Technology.

Ali Ghaffaari: ali.ghaffaari@uni-bielefeld.de

Genomic data must be shared to advance research and medicine, while protecting the privacy of the people it comes from. This seminar covers privacy in genomics in general, with pangenomics as its newest frontier. Each student presents and writes about one research paper. Scientific presentation and writing are learning goals alongside the subject matter.

• Four introductory lectures, then four weeks of preparation, then student presents (up to three per session).
• Attendance is compulsory. If you cannot attend, email me with a reason.
• Papers are chosen from a list we provide: one student per paper, first come, first served. Afterwards, each student picks one other presented paper to act as its discussant.

• Talk: 20 minutes on your paper, followed by discussion and Q&A that you lead.
• Discussant: read one other student's paper and open the discussion of that talk (e.g. two questions and/or one critical point).
• Report: 4,000–5,000 words of main text (about 8–10 pages) on your paper. Any tool may be used; LaTeX is recommended and a template will be provided. You submit a first version, get feedback, and then have two weeks to revise. The revised version is graded.
• Upload your slides to Moodle by the evening before your talk.

The discussant role counts as 1 of the 4 points in the talk's Q&A criterion. Classmates give anonymous peer feedback on each talk; it goes only to the presenter and is not graded.

This seminar follows the yellow tier of the Faculty of Technology's traffic-light policy on AI use. In short:

• Allowed: help with understanding concepts and terminology, translation, suggesting papers to look at, slide layout and design, language editing of your own text, feedback on your own draft, LaTeX and formatting.
• Not allowed: generating slide content, report text, arguments or the outline; generating your discussant questions; citing anything you have not read and verified.
• Declare your AI use in one short statement on your last slide and at the end of your report. Permitted, declared use costs no points.

Details are given in the first session and on Moodle.

• This page: schedule, rules, paper list, templates.
• eKVV mailing list: announcements.
• Moodle: materials, optional forum, submission of slides and reports.
• Email: individual questions to the lecturer.