SCIENCE

The greatest challenge in cancer research is to identify genes that can predict the progress of the disease. The number of genes and gene combinations is vast, so powerful computing tools are required to achieve this. Assisted by Techila Technologies, a Finnish company specialising in autonomic high-performance and technical computing, a research team at the University of Helsinki utilized the huge computing capacity of the Windows Azure cloud computing service for the first time. A computing task that usually requires some 15 years can now be completed in less than five days.

In the treatment of breast cancer, it is of vital importance to be able to predict the likelihood of the development of new metastases. The prognosis for a breast cancer treatment in general is good. ´However, metastases make breast cancer a fatal disease. Cells from the primary tumour infiltrate other parts of the body, forming metastases. The problem is that we are far from understanding this process in sufficient details. We do not know how and which cells become separated from the tumour, where they go or how they act there,' says Sampsa Hautaniemi, docent at the University of Helsinki.

The research team has hundreds of breast cancer tumours at their disposal. The aim is to determine who is most likely to develop metastases by studying the activity of the genes and combining research data.

According to Hautaniemi, computing already plays a significant role in modern medical research because the data available to researchers are getting more extensive all the time. 'It certainly makes the future seem brighter knowing that thanks to cloud computing services, computational resources no longer restrict the setting of research objectives. Earlier on we had to define our research plans based on whether certain computations could be performed.'

Efficiency almost 100 per cent

According to Rainer Wehkamp, founder of Techila Technologies, this was the first time the Windows Azure cloud service was utilised on such a massive scale. 'Our aim was to demonstrate to the University of Helsinki that the technology developed by Techila makes it easy to utilize massive computing power from a cloud. There are also many potential applications outside universities; this kind of computing power could be utilised by e.g. insurance companies in their risk assessment calculations. And as the systems are global in scale, the possibilities are almost limitless.'

The computing was performed in a cloud using the Autonomic Computing technology developed by Techila, and a total of 1,200 processor cores were involved in the operation. The Techila technology was originally developed to utilise the computing power available in table-top computers and servers in businesses and universities, so the transition from a complex multiplatform environment to a homogenous cloud environment presented no problems. 'Azure can be harnessed solely for computing purposes, so its efficiency is extremely high, even as high as 99 per cent,' says Wehkamp.

Microsoft is happy with the results of the pilot project. 'Together, Techila's Grid solution and Windows Azure create a well-functioning and easy-to-use system. Cloud computing takes projects involving high-performance computing to a totally new level. Azure is a massively scalable computing platform service with sufficient resources to complete even the most demanding projects,' promises solution sales manager Matti Antila of Microsoft Finland.