PT  - JOURNAL ARTICLE
AU  - Bjørn Fjukstad
AU  - Vanessa Dumeaux
AU  - Karina Standahl Olsen
AU  - Michael Hallet
AU  - Eiliv Lund
AU  - Lars Ailo Bongo
TI  - Building applications for interactive data exploration in systems biology
AID  - 10.1101/141630
DP  - 2017 Jan 01
TA  - bioRxiv
PG  - 141630
4099  - http://biorxiv.org/content/early/2017/05/24/141630.short
4100  - http://biorxiv.org/content/early/2017/05/24/141630.full
AB  - As the systems biology community generates and collects data at an unprecedented rate, there is a growing need for interactive data exploration tools to explore the datasets. These tools need to combine advanced statistical analyses, relevant knowledge from biological databases, and interactive visualizations in an application with clear user interfaces. To answer specific research questions tools must provide specialized user interfaces and visualizations. While these are application-specific, the underlying components of a data analysis tool can be shared and reused later. Application developers can therefore compose applications of reusable services rather than implementing a single monolithic application from the ground up for each project.Our approach for developing data exploration applications in systems biology builds on the microservice architecture. Microservice architectures separates an application into smaller components that communicate using language-agnostic protocols. We show that this design is suitable in bioinformatics applications where applications often use different tools, written in different languages, by different research groups. Packaging each service in a software container enables re-use and sharing of key components between applications, reducing development, deployment, and maintenance time.We demonstrate the viability of our approach through a web application, MIxT blood-tumor, for exploring and comparing transcriptional profiles from blood and tumor samples in breast cancer patients. The application integrates advanced statistical software, up-to-date information from biological databases, and modern data visualization libraries.The web application for exploring transcriptional profiles, MIxT, is online at mixt-blood-tumor.bci.mcgill.ca and open-sourced at github.com/fjukstad/mixt. Packages to build the supporting microservices are open-sourced as a part of Kvik at github.com/fjukstad/kvik.