
In the Bioinformatics industry, next generation sequencing applications are ideal for cloud migration due to their high performance computing needs as well as the data sharing requirements. However, the savings from cloud are typically lost due to inefficient use of cloud services. A similar trend can be seen with folks attempting to migrate GIS applications to the cloud as well.
The three key concepts for such applications to really save money in the cloud are workload partitioning, dynamic provider selection, and third party security integration.
Workload Partitioning
Bioinformatics applications shouldn’t have to move their Baseline Analyses to the cloud. These are routine applications with very stable usage and should be run in-house with existing resources. On the other hand, their Scientific Analyses are ideal for on-demand high performance cloud computing due to the sporadic nature of this workload. Such workloads also demand very high SLAs from cloud providers. The storage for such workloads should be placed in Network Attached Storage (NAS) devices to facilitate clustering. Cloud storage may not be the best option for such workloads.
Dynamic Provider Selection
Since provider prices and offerings are constantly changing, the best cost provider now may not have the best cost a month from now. Cloud provider performance also varies over time. Therefore, the best cost provider (subject to performance constraints) should be selected at runtime.
Third Party Security Integration
Data in Bioinformatics applications as well as GIS applications are extremely confidential and require a very high level of security. While cloud providers constantly claim increasing levels of security, it is important to integrate 3rd party services that specialize in handling data privacy in addition to data security.
Therefore, Bioinformatics applications, like GIS applications, will benefit significantly from the cloud only if these issues are addressed. It may seem farfetched to expect consumers to figure all this out, but cloud broker platforms have already implemented some of these concepts today. For more information, please visit www.gravitant.com.










