Improving the Performance of Biological Data Analysis in Cloud Computing Platforms

The adoption of distributed database architectures for processing large data sets has shown to be an effective approach to reduce the response time of data analysis procedures. The same approach may be adopted on biological databases, which comprise vast amounts of data that are analyzed to detect genetic diseases, to understand their causing factors and to design process-blocking substances. Defining how data will be fragmented over several network nodes, though, is a non-trivial task. In this work, we have applied a recently proposed methodology for distributed data allocation to create several topology scenarios in a cloud computing infrastructure aiming to improve the performance of data analysis procedures. The proposed scenarios were evaluated using modMine, which is an instance of the Intermine Data warehouse, as case study.