Interview Joris Borgdorff
From AIMSWiki
Q: How did you hear about Africa@home and ultimately get involved in it?
I was working at SACEMA (http://www0.sun.ac.za/sacema/) on HIVMM which is a micro simulation model developed together with Carel Pretorius. HIVMM is a cost-effectiveness tool modelling mainly HIV and tuberculosis and interventions targeting those diseases. I did an internship there for six months, in Stellenbosch, South Africa. HIVMM is a new model, and we really started developing and validating it based on an existing prototype. Shortly after I arrived, SACEMA organized a conference on micro simulation. Roel Bakker from Rotterdam was one of the participants. He had a model that was very similar to ours and which was called STDSIM (a decision support model for control of HIV and other Sexually Transmitted Diseases). So, after the workshop, he came back again to do a more hands-on practical collaboration, to see if the Erasmus Medical Centre of Rotterdam (http://www.erasmusmc.nl/content/englishindex.htm) and SACEMA should work together more closely. It was almost at the end of my internship at SACEMA. I liked micro-simulation modelling and asked him if he had a job for me in Holland. He was the one who told me about Africa@home.
Q: So you started working with the STDSIM team in Rotterdam. Tell me, what is STDSIM all about?
STDSIM models Sexually Transmitted Diseases (STDs). It maps a big network of individuals, tracking how they spread STDs among themselves. The model itself is quite abstract. It tracks different factors, such as for example sexual relationships during sickness, response to treatment, and the impact of related diseases on the evolution of the pandemic. You can use this data to do cost-effectiveness studies to see what kind of interventions have a good impact and why. For instance, the London School of Hygiene & Tropical Medicine (http://www.lshtm.ac.uk) (LSHTM) studied STD evolution in four African cities. They looked at the outbreak and spread records.
Q: And what is your role within the STDSIM team?
I have been trying to do a scale up of the STDSIM model. And part of this has been done. It is now possible to create high volumes of scenarios. The analysis of such a great volume of data however still remains a challenge. I closely work with Roel Bakker, the main developer of STDSIM. He knows the internals. I applied BOINC to STDSIM. There were a few things needed for that to happen. For example, we needed to use Java. This was taken up by another volunteer, Ranaivo, working at the University of Geneva.
Q: Why use BOINC?
The main point is precisely to be able to obtain a lot of computing power. It opens a whole new area of model validation. You can run several models and compare them. Without BOINC, you would not be able to do that. One aspect for STDSIM has also been the positive impact on its public image. It has become more well-known thanks to BOINC. The purpose there has also been to illustrate to some scientists the benefits of micro-simulations models rather than simplistic models.
Q: How do you see your future in relation to the Africa@home project?
In the case of STDSIM, I would like to share some of the knowledge I have, share it with others. Show them how the model works and how it can be used. Also, I would like to remain on standby. I certainly will help the model go live on BOINC. I would also make contact with another group using a method of optimization that deals more with treating large quantities of data. I would like to see what they can do with BOINC and how they would see it move forward. However, I would not be involved in the main use of BOINC.
--Viola Krebs, ICVolunteers 09:30, 17 Jul 2007 (SAST)
