A vision for a biomedical cloud

Abstract

We present a vision for a Biomedical Cloud that draws on progress in the fields of Genomics, Systems Biology and biomedical data mining. The successful fusion of these areas will combine the use of biomarkers, genetic variants, and environmental variables to build predictive models that will drastically increase the specificity and timeliness of diagnosis for a wide range of common diseases, whilst delivering accurate predictions about the efficacy of treatment options. However, the amount of data being generated by each of these fields is staggering, as is the task of managing and analysing it. Adequate computing infrastructure needs to be developed to assemble, manage and mine the enormous and rapidly growing corpus of 'omics' data along with clinical information. We have now arrived at an intersection point between genome technology, cloud computing and biological data mining. This intersection point provides a launch pad for developing a globally applicable cloud computing platform capable of supporting a new paradigm of data intensive, cloud-enabled predictive medicine.

Fig. 1

Projected growth of DNA sequence data in the 21st century.

Fig.2

Biomedical Clouds will interact wit hmedical centres and hospitals providing personalized medicine, as well as researchers and the general public. The blue cloud represents publicly accessible data, whilst pink is used for data that are controlled and have limited access, because of the presence of protected health information, whole human genomes and similar data. Data production may come from many sources, including private and public. There will be multiple Biomedical Clouds associated with medical centers and hospitals that contain controlled data, and these private Biomedical Clouds will be able to ingest data from public Biomedical Clouds. Private clouds may also provide data for research studies via de-identified data sets to public clouds and controlled data to clouds designed to hold such data.