160 likes | 284 Views
Cloud Services for Projects in Bioinformatics : Technical Considerations and Business. Fernando Barraza Omicsco Universidad de San Buenaventura Cali. 10 de abril de 2014. Agenda. What is a Cloud Computing? Main Architecture of the Cloud and the Bioinformatic Business Aspects
E N D
Cloud ServicesforProjectsin Bioinformatics: TechnicalConsiderationsand Business Fernando Barraza Omicsco Universidad de San Buenaventura Cali 10 de abril de 2014
Agenda • Whatis a Cloud Computing? • MainArchitecture of theCloud and theBioinformatic • Business Aspects • A BioinformaticPlatformonthe Cloud • Concernsand Opportunities 10 de abril de 2014
MainArchitecture of the Cloud 10 de abril de 2014
Imagedby Prof. ZHANG Zhang 10 de abril de 2014
Business Aspects • WhereistheMarket? • What are userneeds? • Whatismybusinessmodel? • Who are theplayers? 10 de abril de 2014
BioinformaticsStudies 10 de abril de 2014 Source: Eagle Genomics
How are BioinformaticsStudiesDelivered? 10 de abril de 2014 Source: Eagle Genomics
Perceptionsonbioinformatics Tools 10 de abril de 2014 Source: Eagle Genomics
Cloud Requires Architectural Shift single instance – multi tenancy 10 de abril de 2014
Product vs. Service 10 de abril de 2014
Who are the PaaS players Bitnami Cloud Amazon Elastic Beanstalk 10 de abril de 2014 Slide by SriniKumar, VP MSat
An essential Bioinformatics Cloud Platform Public & Private Online Databases (EMBL, Uniprot, Genomes, etc.) Biomaterial Other Local Databases Databases Bioinformatics and other Tools Browsing Metadata Executing Pipelines (Galaxy, etc) LIMS Searching, Browsing, Annotation (Blast, Clustal, etc.) User Data Files CMS Software Wet Lab Persistent Data Adquisition and Generation Processing, Visualization and Analysis Storage and Management 10 de abril de 2014 Soure: Imaged by Omicsco for Gebix Project
Bioinformatics in Cloud - Maturity Model • Started with small and scripted applications experimenting with Cloud Services • Copy-Paste and Data Files manipulation • Use of predefined Tools and Pipelines • Move to a Hybrid model, where the Cloud Services will integrate with Data Sets in-house • Linked with Public Biological Databases • Enhanced Pipelines • Streamlining of pipelines in a Multi-Instance Environment. • Use of biocomputing ‘appliances’ • Persistent and Large Ubiquitous Storage 10 de abril de 2014
Expectedfeatureson a bioinformaticsappliances • Availability of scientificvalidation of thetoolsthroughpublication/peer review • Computationalefficiency and scalability • Ease of maintenance and installation • Ease of data manipulation/visualization (Do notdiscardcommand-line tool !! ) • Plug & Play forWorkflowPlatforms (APIs and Standard) 10 de abril de 2014
Concerns and Opportunities • Lack of Control, Security and Data Privacy • Service Level Agreements & Standards Compliance • New application model costs/adoption • Data Manipulation still is a pain !! (Web Semantics technologies ?) • Needs for new algorithms to improve performance (big data ?) 10 de abril de 2014
Questions ? 10 de abril de 2014