In addressing some problems reported with using the RefNet package last week, I promised to follow up with an account of the RefNet/PSICQUIC shiny app.
For background: the Bioc RefNet package modestly extends the PSICQUIC package ("Proteomics Standards Initiative Query Interface"). Together they offer annotated interactions from many sources; RefNet adds some hand-curated interaction datasets stored in the AnnotationHub. With very few exceptions, each of the > 150 million searchable interactions includes
- Interaction type
- Detection method
- PubMed Id of article documenting the interaction
- Many alternate identifiers, with protein identifiers dominating
An IDMapper class provides HUGO gene symbols for most of the several kinds of protein identifiers returned by PSICQUIC. RefNet currently adds a human metabolic network (recon2), stamlab TFs, and a 2006 angiogenesis-related hypoxia signaling network; these are intended as examples of the variety of interactions supported. More additions from the community will be eagerly received!
A gentle introduction to the PSCIQUIC/RefNet packages is provided by a Shiny app included inRefNet, which you can try out as demonstrated below. Note, however, that this app is new and may not yet be as robust as an established web app. Clickable links to gene descriptions and PubMed abstracts are provided for all the interactions returned from your search.
library(shiny) runApp(system.file("apps", package="RefNet"), port=9999) Listening on http://127.0.0.1:9999  initializing PSICQUIC...  initializing RefNet from AnnotationHub...  RefNet ready. checking for biomart access... does ' http://www.biomart.org ' respond? creating ensembl mart hsapiens_gene_ensembl dataset provided? connecting to biomart...
Bug reports, questions and suggestions are welcome.