Figure 1

Infrastructure, main tasks, and architecture. A) Infrastructure: Sequencing is performed on an Illumina HiSeq2000 instrument. Data are stored on an Isilon mass storage device. Data are elaborated on a Sun Grid Engine High Performance Computing cluster (SGE-HPC). Application servers run web applications for Genome browsing, data listings, the SMITH LIMS, and host the MySQL information tier. The user data directories are organized by group leader name, user login name, file-type, and run date. B) Sample tracking in SMITH. A sample passes through four states ("requested", "queued", "confirmed", "analysed"). Submitted samples have status "requested". When a sample is added to the virtual flow cell, its status changes to "queued". Upon the group leader confirmation the status changes to "confirmed". The sample is then run and analysed by the workflow engine and assumes the final status "analysed". HPC refers to a high performance computing cluster. C) Architecture of the workflow unit. Generated commands invoke Galaxy workflows that subsequently call the un-pluggable core. A part of the instruments can be on the Galaxy side (proprietary tools and scripts) and the other part (open-source tools) is moved to the core.