Gene Cloud - Development of novel methods based on Steganography for scalable data protection and development of novel drug screening algorithms for a cloud environment

The pharmaceutical industry faces ever growing costs for the development and validation of new drugs, while the rate of new drug approvals is constant. Increasing amounts of data are or will be available, comprising experimental or patient data and data from computationally demanding large scale bioinformatics computations or simulations. Cloud Computing models offer such needed computing power. However, pharmaceutical data have a high privacy demand, since they usually contain valuable findings and underlie confidentiality regulations.
As data privacy issues are a major concern for companies regarding the use of publicly available Cloud solutions, several proposals were made in the recent years. These include certification of the Cloud providers and the underlying data centers, introduction of co-processors to the Cloud systems (trusted computations via trusted platform modules), and the use of fully homomorphic encryption (rigorous data security). However, these solutions are either rather ineffective at supporting data privacy, or inefficient, and therefore will not scale to typically used amounts of data. Furthermore, these solutions cannot easily be adopted to achieve a viable trade-of between data protection and efficiency.

To offer secure Cloud Computing services especially to smaller enterprises, GeneCloud develops Cloud services based on a data-centric security concept. This involves data anonymization, blinding and/or encryption to realize flexible and scalable data security.

PartnerS

• TU Dresden
  • BIOTEC
  • Database Technologies
  • ZIH
• Transinsight GmbH
• Qualiype AG
• antibodies-online GmbH

Project Term

11/2011 - 10/2014

FUNDING

Bundesministerium für Wirtschaft und Energie (BMWE)

Publications

• M. Beck, V.J. Haupt, J. Roy, J. Moennich, R. Jäkel, M. Schroeder, Z. Isik: GeneCloud: Secure Cloud Computing for Biomedical Research, Lecture Notes in Computer Science, 2013.
• J. Roy, C. Winter, Z. Isik, M. Schroeder: Network information improves cancer outcome prediction, Brief Bioinf 2012; doi: 10.1093/bib/bbs083
• M. Beck, F. Kerschbaum: Approximate two-party privacy-preserving string matching with linear complexity, Computing Research Repository (CoRR), September 2012.