What is

The project is an application that makes use of network computing for stochastic modelling of the clinical epidemiology and natural history of Plasmodium falciparum malaria.

Simulation modeling of malaria

The fight against malaria was given a new impetus by the call for eradication at the Gates Malaria Forum in October 2007, making more but still limited resources available for research, development, and combating malaria. To inform decisions on which new or existing tools to prioritize, we have developed a general platform for comparing, fitting, and evaluating stochastic simulation models of Plasmodium falciparum malaria, programmed in C++ ().

We use this to inform the target product profiles for novel interventions like vaccines, addressing questions such as minimal efficacy and duration of effects needed for a vaccine to be worthwhile, and also to optimize deployment of established interventions and integrated strategies. Field trials of interventions consider effects over 1-2 years at most, but the dynamics of immunity and human demography also lead to longer term effects. We consider many different outcomes, including transmission reduction or interruption, illness, hospitalization, or death, as well as economic aspects.

Malaria occurs in an enormous variety of ecological settings, and interventions are not always universally applicable. For instance, indoor residual spraying works only with indoor-resting mosquitoes, and insecticide treated mosquito nets only with nocturnal vectors. The best combinations of interventions vary, as do optimal delivery approaches and their health system implications. There are trade-offs between high coverage and costs or feasibility of deployment. Indiscriminate deployment may lead to evolution of drug resistance or insensitivity to other interventions. To support the analysis of these elements we are assembling databases of health system descriptions, intervention costing, and vector bionomics across different malaria ecotypes.

Uncertainties inherent in simulations of complex systems are addressed using probabilistic sensitivity analyses, fitting multiple different models, and basing predictions on model ensembles not single simulations. This requires super-computing, both for statistical fitting (which must simultaneously reproduce a wide range of outcomes across different settings), and for exploring predictions. We obtain this computing power over the internet from spare capacity on the computers of volunteers (

Meetings with potential users of these predicitons are used to promote the models and their predictions to wider communities of malariologists, planners, and policy specialists. We are also developing web-based job submission and analysis systems to increase internet access to models.

This project is supported by Bill and Melinda Gates foundation logo

Powered by BOINC Logo

Current Research team

Applied Mathematics

  • Melissa Penny (Swiss TPH)
  • Nakul Chitnis (Swiss TPH/MACEPA)

Epidemiology/Public Health

  • Allan Schapira (Swiss TPH)
  • Don de Savigny (Swiss TPH)
  • Marcel Tanner (Swiss TPH)

Quantitative Biology

  • Ian Hastings (LSTM)
  • Katherine Winter (LSTM)
  • Olivier Briet (Swiss TPH)


  • Konstantina Boutsika (Swiss TPH)


  • Tom Smith (Swiss TPH)
  • Amanda Ross (Swiss TPH)

Computer Science

  • Diggory Hardy (Swiss TPH)
  • Aurelio di Pasquale (Swiss TPH)
  • Nicolas Maire (Swiss TPH)
  • Michael Tarantino (Swiss TPH)
  • Henning Mortveit (VBI)

Health Economics

  • Fabrizio Tediosi (Milan)
  • Josh Yukich (Tulane)
  • Valerie Crowell (Swiss TPH)
  • Lesong Conteh (LSHTM)
Contact show/hide

Return to main page

Copyright © 2013 africa@home