Breast cancer is the most common cancer type for women. While the prognosis of many subtypes of breast cancers in general is good, certain types of breast cancer pose high risk for metastases, which account for 95% of the deaths of breast cancer patients. In the treatment of breast cancer, it’s important to be able to assess the likelihood that a primary tumor develops metastases and this is a non-trivial task, as there’s an ocean of genes and their combinations that may contribute to metastasis progression.
For the first time, a research team at the University of Helsinki has utilized the enormous computing capacity of Windows Azure with the help of Finnish HPC solutions provider, Techila, to dramatically speed genetic research that will improve breast cancer treatment. As University of Helsinki docent Sampsa Hautaniemi explains, “currently, many breast cancer patients are treated with aggressive regimens. When we are able to predict which primary tumors are more prone to result metastasis, we can suggest the level of treatment on a more individual basis already at the time of diagnosis.”
The greatest challenge the research project faces is finding just the right combination of genes from among 25,000 genes. There are an enormous number of different possibilities, which means that high computational power is required. Windows Azure has dramatically speeded up the process by enabling researchers to complete computational tasks in four and a half days that would normally take around 15 years to complete. Rainer Wehkamp from Techila is pleased that the pilot project specifically concerned cancer research. “One in four people in the world gets cancer and although we do not possess the skills to carry out the actual research, it’s nice to be able to contribute to it through our own expertise.”
Computation was performed in a cloud with the help of the autonomic computing technology developed by Techila, using 1,200 processor cores. As Hautaniemi explains, “These kinds of pilot projects are important, as computation is a very essential part of modern medical research, and the data sets are growing all the time. The fact that computational resources provided by the cloud are available increases prospects to conduct really demanding projects that were not possible few years ago.”
Notes, Wehkamp, “An average of 95 % of the capacity is wasted at universities. Windows Azure harnessed for computational use has a high utilization rate of up to 99%. The speed, efficiency and ease of use of the computation are competitive edges in both the research and business worlds.”