Fig 1.
Overview of STENCIL architecture.
(A) A ReactJS frontend server provides web-access to the data. A NodeJS server and MongoDB instance store and manage the data provided and disseminated through RESTful API calls. Data analysis and hosted URLs are provided by the Galaxy platform and communicated directly to the MongoDB. Local analysis and file-hosting outside of the Galaxy platform is also supported. (B) A sample React-served experiment section demonstrating multiple types of data. React serves static PNG/JPG/SVG images hosted remotely. Interactive charts include dynamic mouseover with additional data, the ability to expand to full screen, and export to SVG. (C) Data table analysis allows for sorting and filtering using remotely hosted data. (D) Integrated Genome Browser (igv) provides interactive track visualization in the same web frame as other advanced analysis.
Fig 2.
STENCIL backend enables precise project organization and authorization (A) MongoDB model schema of STENCIL backend.
URL references to experiment analysis are stored in the Experiment model. Experiments are linked to the Project model through a common projectId string. Users gain access to models and underlying experimental data through assignment to a variable number of projectIds. (B) Experiment navigation screen of STENCIL. STENCIL supports users accessing multiple public and private projects as defined by the backend database. Real-time search enables quick filtering and access of experiments in large projects. (C) STENCIL Admin console is available to STENCIL administrators and allows for project public/private control and project summary edits. (D) STENCIL Admin console provides control of user access to projects and defining user roles.
Fig 3.
Galaxy integration with STENCIL.
(A) A custom python script provides the mechanism through which a Galaxy tool (i.e., DESeq2) can POST its output to STENCIL (B) The STENCIL-Galaxy communication tool allows the user to specify the type of data being transmitted to STENCIL (C) A sample JSON payload containing the minimum amount of information needed by STENCIL to correctly place the analysis within an experiment and associated project.
Fig 4.
Application of STENCIL in the PCRP was an NIH-initiated project to screen 800 antibodies in ChIP-exo. NIH project requirements included providing the generated data in a community-accessible manner.
Fig 5.
Application of STENCIL in the Yeast Epigenome Project.
A comprehensive map of protein binding in S. cerevisiae required the development of a Galaxy workflow to both generate quality control metrics as well as provide baseline biological insight. Full analysis with high-resolution figures is available at http://www.yeastepigenome.org/yep/factor/ABF1.
Fig 6.
DESeq2 RNA-seq analysis visualized in STENCIL.
(A) DESeq2 charts are visualized as dynamic charts in STENCIL. All data is generated and hosted directly from Galaxy. (B) The nivo charting library allows for on-the-fly generated of interactive plots containing tens of thousands of unique datapoint in seconds.
Table 1.
Evaluation of STENCIL load times.