Course: Optimisation Techniques


May 17-21, 2010, University of Amsterdam, Amsterdam


Dr. Jaap Kaandorp
Prof. dr. Antoine  van Kampen
prof. dr. Jaap Heringa
Contact:  Dr. Jaap Kaandorp

Study load

The study load of this course is 3 EC. Participants will get a certificate after successfully completing this course.


Room F3.20, Science Park 107, 1098 XG Amsterdam


After this course a student should have an overview and basic understanding of optimisation methods frequently applied in bioinformatics and systems biology.

Target audience

The course is aimed at PhD students with a background in bioinformatics, systems biology, computer science or a related field. A working knowledge of basic statistics, linear algebra and differential equations is assumed but will be reviewed during the first day of the course.


Optimization, in general, is concerned with finding one or more optimal solutions given a problem. Many optimization problems are very difficult to solve. In many different problems from bioinformatics and systems biology (e.g. multi parameter estimation, reverse engineering of gene networks, multi-alignment problem, 3D structure prediction etc.) various optimisation methods are applied.

In this course you will get acquainted with the underlying mathematics of optimization and with a selection of local and global optimization methods. In addition, several examples of optimization problems in life sciences will be presented and discussed.

We will compare methods like linear programming, steepest descent and conjugate gradient. We will discuss global optimisation methods like Monte-Carlo sampling, the Basin hopping techniques (aka Monte-Carlo with minimization – e.g. ICM (internal co-ordinates system for protein 3D structure prediction),  simulated annealing and evolutionary algorithms. Applications of stochastic optimisation  in systems biology, hybrid methods using stochastic optimisation in combination with local search will be discussed.

Examples will that will be discussed during the course include: (a) multi-sequence alignment with simulated annealing and evolutionary algorithms and comparison to dynamic programming, (b) parameter estimation in models of large biochemical networks and the application in reverse engineering of spatio-temporal models of gene regulation.

Tentative course programme

The course will consist of lectures, computer practicals, and furthermore on several course days there will be  paper readings & paper presentations in groups.

Date Topics
May 17 Introduction and Foundations

  • Foundations of optimisation: the mathematics
  • Introduction local univariate optimisation techniques
 May 18  Local Optimisation

  • Local optimisation techniques: Simplex method, Steepest Descent, Conjugate gradient, and others
  • Computer practical: Simplex (Nelder-Mead) and related techniques
May 19 Stochastic Optimisation

  • Stochastic optimisation techniques: Monte Carlo, Simulated Annealing, Particle Swarm Optimisation
  • Computer practical: Simulated annealing and Particle Swarm Optimisation
May 20 Global Optimisation

  • Global optimisation techniques: Genetic Algorithm, Dynamic programming, Expectation Maximisation
  • Computer practical: Genetic Algorithm, Dynamic programming
May 21 Applying Multi-Objective Optimisation

  • Mixed and hybrid techniques: Application to Genetic networks


The next course is expected in 2013.