Chemical Reaction Network Simulation

Reaction Network Simulation

This course teaches simulation methods for reaction networks, for example networks of chemical reactions, such as gene regulation networks. The networks are not spatially resolved. We discuss how to represent networks in the computer at different levels of detail, and what numerical simulation algorithms exist for dynamic processes on networks. We place particular emphasis on algorithmic efficiency and on efficient data structures that allow reducing the runtime complexity of the resulting simulations.


Contents

stochastic reaction network descriptions, deterministic reaction network descriptions, dependency graphs and network coupling analysis, exact stochastic simulation algorithms for chemical reaction networks, approximate stochastic simulation algorithms for chemical reaction networks, deterministic simulation algorithms, efficient data structures for network simulations.


Time/Place

Lecture: Mondays, 4. DS (13:00-14:30) in APB-2026 (computer science building) / FIRST LECTURE: OCT 24
Exercises: upon agreement


Teachers

Lecture: Prof. Ivo F. Sbalzarini
Exercises: Dr. Benjamin Dalton


Learning goals
  • Graph-based network representation (reaction graphs vs. dependency graphs)

  • Network descriptions at different levels of detail, from exact stochastic to approximate deterministic

  • State-of-the-art simulation algorithms for chemical reaction networks

  • Complexity analysis of the algorithms and dependence on data structures used

Special remarks

We focus on chemical reaction networks in biological systems. The taught methods and concepts are, however, applicable in a much broader sense. Full lecture notes will be provided to the students of the course.


Lecture language: ENGLISH


Please find below the lecture syllabus and handouts:
  • Lecture 1 - Administration and Introduction; what are reaction networks and where are they used?
  • Lecture 2 - Graph representations (reaction graph vs. dependency graph), classification of networks, coupling degree
  • Lecture 3 - Stochastic chemical kinetics and the classic SSA algorithm
  • Lecture 4 - Algorithmic complexity of SSA and more efficient formulations
  • Lecture 5 - Partial propensity methods and efficient data structures
  • Lecture 6 - Approximate stochastic chemical kinetics and approximate SSA
  • Lecture 7 - Deterministic chemical kinetics and deterministic simulations
  • Lecture 8 - Time stepping and stability in deterministic simulations
  • Lecture 9 - Comparison of deterministic and stochastic simulations: when to use which?
  • Lecture 10 - Simulating reaction networks with time delays: priority queues
  • Lecture 11 - Simulating spatiotemporal systems: Next-Subvolume Algorithms
  • Lecture 12 - Identifying network parameters from data
  • Lecture 13 - Q & A / backup
  • Lecture 14 - Presentation of student projects