BioStatsLab

BioStatsLab (a BSc Bioinformatics Replacement for MInf3)

Basic mandatory course, B.Sc. Bioinformatics, Saarland University.
This course is not available for credit points to students of other programs! You are welcome to audit, but will not get ECTS.

Prerequisites	Mathematics (MInf1+Minf2, especially some  analysis and linear algebra); solid programming skills
Credits	9 ECTS credits
Required time	4V+2Ü (4 hours of lectures, 2 hours of tutorials per week)
Language	English (although this is a basic course, it also serves as an additional prerequisite for international Master students, and is hence given in English!)
Registration	click on Registration in the menu header
Materials	Materials will be available after registration under Information > Materials
Times	Lecture: Wednesday 08:30 - 10:00 and Friday 12:15 - 13:45. Tutorials: Wednesday, 14:00–16:00, Thursday, 08:00–10:00 Office Hour: TBA
Mode	lecture in presence in E2.1, room 001
Link	https://cms.sic.saarland/biostatslab25
Instructor	Prof. Dr. Sven Rahmann
Tutorials	M. Sc. Inês Alves Ferreira
Exam	Written exam (up to 3h) at the end of the semester. Requirements to participate in the exam: - No score requirements for the exercises (pointless since LLMs can solve all of these) - Presentation of your solved exercises before the tutorial group (as explained in the first lecture)

 

Target audience 

This course is offered as a basic lecture in the B.Sc. Bioinformatics program as a replacement for Mathematics for Informaticians 3 (MInf3).
Thus it should be taken in the 3rd semester, after completing MInf1 and MInf2, as well as Programming 1 & 2.
It should be taken in parallel to Bioinformatics 1 during the B.Sc. Bioinformatics program.

You will need some programming skills to qualify for the exam. Best would be Python, but you can use a language of your choice.
Please do not waste your time by attempting this course without a solid basis in programming.

 

Topics

The following topics will be covered in the course; additional topics may be included, depending on time and current events.

Probability

• randomness
• uniform distributions on finite sets (Laplace spaces)
• elementary and advanced combinatorics
• finite, discrete and continuous probability spaces
• random variables
• discrete probability distributions and how they are derived
• probability distributions and OOP, scipy.stats
• conditional probabilities
• Bayes’ Theorem, simple version
• moments of random variables (expectation, variance, …)
• continuous probability distributions
• a glimpse at measure theory
• posterior distributions

Statistics

• descriptive statistics
• parametric models
• statistical testing (frequentist view)
• statistical testing (Bayesian view)
• parameter estimation: moments, maximum likelihood
• parameter estimation in mixture models: EM algorithm
• regularization and Bayesian view on estimation
• linear regression
• robust regression
• multiple regression
• logistic regression

Stochastic Processes

• stochastic processes
• models for random sequences
• Markov chains
• Markov processes: models of sequence evolution
• Hidden Markov Models and applications
• Probabilistic Arthimetic Automata (PAAs) and applications
• the Poisson process
• distribution of DNA motif occurrences: compound Poisson
• significance of pairwise sequence alignment

Applications in Bioinformatics

• tests for differential gene expression
• Bayesian view on differential gene expression
• high-dimensionality low-sample problem
• multiple testing

Multi-dimensional analysis (compact, 1 week)

• partial / total differentiability
• high-dimensional optimization