Statistics, Probability, and Algorithms in Bioinformatics (SPAß)

(Algorithmen für Stochastik und Stochastik für Algorithmen)

Lecture and Exercises

New + Important

11.02. In the week before the Klausur, an uncorrected version of the lecture notes sources and the resulting pdf is available.

03.12. Because of the ongoing strike, today's lecture will take place at the MPI for Molecular Genetics, Ihnestr. 73, 3rd floor in the small library room.

26.11. Because of the strike, the lecture could not take place today. The problem sets that were due today can be handed in next week. Also, be prepared for more and more difficult problems next week. You might also start reading about the accept-reject method for random number generation. (Even though there is a strike, there is no reason to learn less.)

19.11. About note-taking: I will say in class how to obtain the most recent version of the lecture notes and the source files. Please do not change any of those files or hand in modified versions. Instead, please send me your notes in a file notes_yourname.tex. It should contain a chapter and any new commands that you might need (but please use the provided commands as far as possible). Look at previous lecture notes for examples.

31.10. In problem 6 of the exercises, the file intronlengths.txt that you can download from the website, contains the lengths of individual introns, not the length distribution! So the mean intron length is approximately 882. Thanks to Jonas, and sorry about the confusion.

28.10. Lectures start each Wednesday at 14:00, not 14:15. They take place in Room 025/026 in the Pi-building, Arnimallee 2-6. The exercises start immediately after the lecture, at 15:30.

16.10. I put the first set of exercises online beforehand, so you know what to expect...

Summary

This lecture is about fun (Spaß): Statistics, Probability, and Algorithms in Bioinformatics. In other words, we'll be looking at the many situations where stochastics and algorithms meet. After this course, the successful participant will be well prepared with a toolbox of methods from both worlds and their applications to sequence analysis.

Scheinkriterien

Vorlesung und Übung können nur zusammen belegt werden. Die beiden Teilnoten ergeben gemittelt die Gesamtnote.

Vorlesung (3 credits):

Ausarbeiten eines Skripts zu einer Vorlesungsstunde mit LaTeX (50% der Note); Bestehen der Klausur am Ende (50% der Note). Beide Noten müssen mindestens ausreichend (4.0) sein.

Übungen (3 credits):

Die Note richtet sich nach der erreichten Gesamtpunktzahl in den Übungsaufgaben. Zum Bestehen sind mindestens 50% zu erreichen. Ein großer Teil der Aufgaben wird aus Projekten bestehen. Eventuell lassen sich aus einzelnen Projekten Themen für Master-Arbeiten ableiten.

Literatur

• William H. Press, Brian P. Flannery, Saul A. Teukolsky und William T. Vetterling
  Numerical Recipes in C, Second Edition
  Cambridge University Press
  A classic.
• John A. Rice
  Mathematical Statistics and Data Analysis, Second Edition
  Duxbury Press
  Relatively expensive, but easy-to-read introductory statistics text.
• W.N. Venables, D.M. Smith, and the R Development Core Team
  An introduction to R (Notes on R: A Programming Environment for Data Analysis and Graphics, Version 1.8.0, 2003-10-08)
  Official web page of the R project: http://www.r-project.org
  R is a free version of the commercial Splus software. Look here for further books on R.
• Christian P. Robert und George Casella
  Monte Carlo Statistical Methods
  Springer
  Not very easy to read. Has a lot of material about random number generation and MCMC methods. Contains quite a number of misprints for a SPringer statistics text.
• Ronald L. Graham, Donald E. Knuth und Oren Patashnik
  Concrete Mathematics
  Addison-Wesley
  Contains all the mathematics a programmer or algorithmician will ever need -- and more. Not very easy material, but extremely well written..
• Robert Sedgewick und Philippe Flajolet
  Analysis of Algorithms
  Addison-Wesley
  Has many examples about how to use generating functions for the analysis of algorithms.

Weitere folgen. In der Vorlesung werde ich jeweils einzelne Themen und Beispiele aus den genannten Büchern herausgreifen.

Zur Vorlesung soll von den Teilnehmern ein Skript erstellt werden. Deswegen wird jeder Teilnehmer mindestens einmal eine Vorlesungsstunde als LaTeX Skript nachbereiten. Dies macht 50% der Vorlesungsnote aus.

Links

Homepage der Abteilung Computational Molecular Biology am Max-Plack-Institut für Molekulare Genetik.

Zum Studiengang Bioinformatik an der FU Berlin.