This graduation project is about optimally designing biomolecular sensors. The sensors we aim at are designed to detect DNA fragments present in samples such as blood samples. It contains a number of sensory elements (sensels), that each test for a small, specific subsequence of DNA. It can be used, for instance, to detect the presence of a certain bacterium, by having sensels that can detect specific DNA subsequences of that bacterium. In this way one can quickly and easily determine the bacterium causing an infection, which can be used to select proper antibiotics. For diseases such as sepsis, this may save many patients’ lives, as it allows for a diagnosis in a few hours, instead of several days for conventional lab tests.
The design of such a biosensor, with a few hundred sensels, should be done with care. The reason for this is that neigboring sensels may influence each others reading, or that one want to have more robust results by replicating sensels, and spreading them over the sensor to reduce the effect of local noise. Furthermore, this replication may be done differently for different sensels (some may be duplicated, others may have to be put on the sensor in four-fold). Additionally, there may be constraints and objectives to be taken into account resulting from the way the sensors are produced.
Hence, the optimisation problem that we want to address is how to place sensels on the sensor, such that the read out of the sensor is as good (reliable) as possible.
The project will consist of a few aspects:
- how to formally model the biosensor design problem,
- a complexity analysis of the problem,
- designing and implementing a solution approach, and
- evaluation the approach, by testing it on a number of instances.
The candidate student should have a background in combinatorial optimization, discrete algorithms, and heuristic search.
Concact: dr.ir. Wim F.J. Verhaegh
Philips Research Laboratories
Prof. Holstlaan 4
5656 AA
040-2744931
wim.verhaegh@philips.com