Algorithmic Bioinformatics (Prof. Canzar)

This seminar covers foundational algorithms and statistical approaches in bioinformatics, as well as current topics such as long-read sequencing, single-cell omics, epigenomics, genome assembly, and gene regulatory networks.

Starting from a text (book chapter, journal article), you will perform an in-depth literature research to grasp the relevant biological background and computational foundations. You will prepare and deliver a scientific presentation on that topic, manage the subsequent Q&A, and actively participate in peer discussions.

Numerous problems in bioinformatics and other domains can be formulated as optimization tasks; we seek the shortest, cheapest, or longest among many possible solutions that satisfy certain constraints. For some problems, there is no known algorithm that can find the optimal solution efficiently, or the problem can be proved NP-hard. Even in such cases, we are interested in a solution that is as good as possible, to learn something about the underlying (biological) problem.
This course introduces fundamental techniques for modeling and solving combinatorial optimization problems. The first half of the course emphasizes exact methods, especially those based on (integer) linear programming. The second half explores heuristic and metaheuristic approaches, which provide practical solutions when exact methods become infeasible.

A meaningful interpretation of biomedical data requires a basic understanding of the biological units and processes from which the data were generated as well as knowledge of the technologies that produced the data. The aim of this module is to equip students with this understanding. It will cover fundamental concepts in molecular and cellular biology, genetics, and strengths and limitations of current DNA sequencing technologies. Furthermore, it will highlight how biological questions can be translated into computational problems, from sequence alignment and genome assembly to gene regulation and disease variant detection.

Computational genomics increasingly depends on complex data analysis pipelines, making reproducibility essential for reliable scientific research. In this seminar, students learn the principles of reproducible genomic data science through a hands-on project: working in teams, they reproduce selected analyses from published genomics studies, identify reproducibility challenges, and develop reproducible workflows using Snakemake while applying best practices for computational research.