Benjamin Schuster-Böckler: Cancer Informatics
Cancer research now generates huge amounts of data, and sophisticated computational tools are needed to answer biological questions. Making sense of this variability at the molecular level will help us better tailor treatments to individual cancer patients.
Q: What is bioinformatics?
Benjamin Schuster-Böckler: Bioinformatics is a fairly new field of research that tries to answer biological questions using computers or computational tools. Over the last 20 years or so there has been a revolution in the biological sciences with new experimental techniques that produce huge amounts of data. For one cancer patient you would produce so much data that it would take you a whole day to download on your home internet. We don't have only one of these data sets, but thousands. It really is almost impossible now to do research without using computer programs to actually sift through that data to find the biologically relevant aspects and distinguish them from random noise.
Q: What are the most important lines of research that have developed in the last five to ten years?
BS: What I find particularly striking that has come out of the genomics revolution in the last five or ten years, is that we are now aware that cancer is an incredibly diverse disease. It's almost as if it is not just one disease but many diseases. Each patient can have very similar presentation in the clinic, but if you look at the molecular details and the genomes, these patients can be incredibly different. It is really becoming clearer that those differences can have a great impact on the treatment success of these patients. This field known as cancerheterogeneity is one of the things that I have found particularly exciting.
Q: How is bioinformatics helping us understand cancer?
BS: Nowadays with the explosion of data that is being generated in cancer research, it is almost not a question of how it can help us understand cancer, but it is almost impossible to do any cancer research without the help of computer tools. To put this into perspective, it is very common nowadays to generate data sets which have millions of rows and tens or thousands of columns, and just by looking at them you would not be able to understand what they mean. We really need computer programs to help us detect the biologically relevant aspects of the result and distinguish them from those aspects which are effectively noise.
Q: Can you tell us about your research?
BS: The research in my group is focussed on trying to understand the origins of variability in cancer. To explain it better I can use an analogy. A mutation in the genome is like a car accident that happens on a stretch of road; but not all roads are equally dangerous. Even though each accident might be bad luck, the frequency of accidents varies massively between different roads. What we are trying to understand is the specific likelihood of an accident is in a particular area on average. The important question is whether there are specific roads where accidents happen with a far greater frequency than on average. Those are the regions that are of particular importance for disease. In the genomic context, there are some regions in the genome which accumulate mutations at a much greater rate than similar other regions in the genome that have similar characteristics and these are the ones that we want to identify.
Q: Why does this research matter and why should we put money into it?
BS: We can see now that patients with cancer are very different on the genomic level. It is crucial for us to understand which ones of the mutations in the patient genomes are driving the cancer, the ones that are important for the development and the progression of the disease, and distinguish them from those variations which are just passengers. In order to be able to distinguish those drivers from the passengers we need to have an idea of what the overall likelihood of mutations is in the genome on average. That will allow us to pinpoint those mutations that happen at much greater frequency and those ones are usually the ones which are relevant in the disease context.
Q: How does your research fit into translational medicine within the Department?
BS: I think it is generally accepted that in the future we will see a lot more personalised medicine. When I say personalised medicine, I mean that we will start tailoring treatment to an individual patient's genomic characteristics. In order to be able to do that we need to understand which aspects of an individual's genome are predictive for treatment success and which ones have absolutely no impact on the disease. So, one of the key aspects of the research that we are doing in our lab is helping to make that distinction.