I agree with Pierre, that a general purpose R GUI like Rstudio or Biocep-R provides a good enough front end for bioconductor. If you want to hook Bioconductor into other bioinformatics systems, you might want to look into efforts to wrap Bioconductor in Galaxy. It looks like this has not come to fruition, but there many notes on the Rgalaxy wiki that may be use if you go down this road.

RStudio ? http://www.rstudio.org/

Red-R is a GUI interface to R that includes some of the Bioconductor packages: http://www.red-r.org/

My favorite is a combination of Eclipse and StatET

Although I cannot point you to a specific reference or link, I remember reading about and surfing on several web interfaces to Bioconductor. However, the functionality of these interfaces was always limited compared to what is available in Bioconductor. This is essentially because these web portal were focused on certain pipelines, technologies, ...

One could even question the usefulness of such a complete interface if it existed, because it would limit the flexibility of the R environment. Bioconductor is not only about the functions/classes/methods that are distributed with the packages, but it is very much about the flexibility of R for data manipulation and exploration.

That said, you might find interesting workflow interfaces googling for R/Bioconductor and Taverna.

There have been several attempts to develop R web interfaces over the years; 2 current working projects are RApache and Rook. There's also this site at the University of Surrey which provides several web interfaces including one for RankProd. However, nothing else specific to Bioconductor so far as I know.

In some ways, Bioconductor is not a good candidate for web development since (1) it's often used in an exploratory fashion, rather than in defined workflows and (2) many of its tools (e.g. normalization) are CPU-intensive, therefore slow, which complicates web applications - they'd require batch queues, email notifications and so on.

There is also affylmGUI which is a Tcl/Tk interface to Limma for Affymetrix expression chip analysis, it also includes QC routines etc.

GEO2R is pretty nice. If you visit the Web page for a gene expression data set in GEO (example: http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE862) you'll see a link labeled "Analyze with GEO2R." If you follow the link, you'll find a Web interface that lets you assign groups and run an analysis using limma. It will show you the results and also the R/Bioconductor code that ran the analysis. I don't think many people know about it ... it's fun to show to people because it gives them an automatic tutorial on how to use some of the tools and libraries.

Has anyone considered using TACTIC as a workflow GUI and engine for Bioconductor and R? I think it would be very possible to integrate R scripts into to TACTIC using a module like rpy2.