How can we use the expertise within our IT field to help other scientific disciplines develop? How can we maximise our use of the full computing power of current computer technology to optimise and speed up processes? That is the focus in Computational Science.
How can we advance other sciences using our knowledge of computer science? That is essentially what Computational Science is all about. Computer technology is used today in every industry and discipline, but it is rarely used to its full potential. As a result, many opportunities are missed.
I studied physics after my master’s degree in computer science, and obtained my doctorate at the interface of the two disciplines. During my studies I noticed that a knowledge of computer science can help enormously with making progress in other sciences. All researchers use a computer, but often they have an inadequate understanding of the possibilities of that machine. As a result, they miss out on a lot of opportunities - in terms of speed, automation and optimisation of processes.
Take the computing power of computers for example: today it is huge. With the right programs, these machines can relieve you of a lot of manual tasks. Computers can perform many calculations better, faster and more efficiently than humans. Their margin of error is lower. Moreover, many processes can be automated relatively easily. There are still considerable time-savings to be made in this area today. And time savings mean more research results that are reliable, and a higher research output.
A while back I was working with the Hasselt University researchers from BIOMED. Their focus was on DNA sequencing. Until recently, they had been using a number of online tools to analyse their data. It was an extremely time-consuming process with many intermediate steps. They first copy-pasted all their research results into an online database. Next they had to wait for the tool to carry out its calculations, because of course they weren’t the only ones using the database. They then had to manually filter out a number of results that needed further investigation. These in turn had to be processed by another online database. It was a particularly mind-numbing and time-intensive task that doctoral students often spent weeks on.
At a certain point they wanted to start using a next generation sequencer: a device that can read huge numbers of DNA sequences at the same time and therefore also produces a vast amount of data. However, they were afraid that they would not be able to cope with this even larger stream of data. At this point they came knocking on EDM’s door. What could be automated? And how could things be organised more efficiently?
Bridging the gap between two disciplines is not easy. You have to learn to speak each other’s language and be able to define very precisely what you want to research, how you will proceed and on the basis of what parameters you will make selections. This requires a serious effort from both parties. As computer scientists we have to completely immerse ourselves in an unfamiliar research field, while they have to learn to split up their research processes clearly and very accurately into thousands of little intermediate steps. Overcoming these barriers is not easy, but it always pays off!
By getting together to thrash things out at regular intervals, we quickly got a much better idea of what our colleagues at BIOMED were asking for. Based on the input they provided, we wrote a first version of a program. We then subjected this to tests that we went through together. What worked well? Where were there still gaps? And for what problems did we still need to find a solution? On the basis of these results, we adapted the program each time and refined it further. In this way, after a few weeks we were able to deliver a relatively simple program that analyses hundreds of sequences in a few minutes. The time-savings achieved with this are enormous. Thanks to the new program, BIOMED can now analyse more data in a shorter time, and study particularly interesting data in more depth. In addition, the program provides a much more reliable analysis of the DNA sequences than a manual analysis. This will of course be of benefit to BIOMED’s research output.
It gives me immense satisfaction to help strengthen other scientific disciplines by sharing our knowledge of computer science with them. That’s what Computational Science is all about for me. We put on our computer science glasses and look at the research processes in other scientific disciplines. We look for opportunities and try to use the full potential of our field. In a spirit of openness to and interest in other scientific research domains, we try to solve the puzzle together. Within the university we have already collaborated with biomedical sciences and biostatistics, and with chemists and physicists. We have already been able to make significant progress in all these areas.
What makes this research field complex? You never know in advance exactly what you can do for the other party. First you have to get to know the new field in depth, explore possibilities and find out which optimisation methods will make a difference in that particular field. There are no standard solutions. Every case is different and requires a tailor-made solution. But that is precisely what makes this domain so fascinating.