Languages and Methods for Query-oriented Data Mining in Bioinformatics

Recent technological developments allow researchers in microbiology to run high-throughput experiments at the cell level, for example using time-lapse microscopy. Such experiments produce a lot of structured data that is not easy to analyse. Data-mining provides many techniques to extract knowledge from data that can help in this context. However applying existing algorithms is not always trivial because of several reasons: the problem formulation between the two fields can differ, the existing systems can require programming skills, or some algorithms can necessitate knowledge of the algorithm itself to be executed correctly.

The goal of this PhD is to bridge this gap between the 2 fields. On one hand, it means identifying data-mining methods that can analyse time-lapse microscopy data and possibly adapt and/or extend these methods to this specific context. On the other hand, it means building systems that can be used by non data-mining experts. To do so, the research is focussed on query-oriented approaches, where a user can query for answers in a declarative way that does not require any knowledge of the method itself.

This PhD will be done in collaborations with microbiologists from the Lab of Food Microbiology of the LandbouwInstituut of the KULeuven.