[Bioc-devel] Interfacing libraries not part of Bioconductor
Raffaele, In order to minimize the problems for the end-users and the build system, the ideal situation would be to have JETTA as part of the Bioconductor project. Have you talked with the JETTA authors to see if they would be interested in joining the Bioconductor community? This could help increase the exposure of their software as well as minimize their long-term effort of maintaining a distribution site and testing system. Patrick
rcaloger wrote:
Hi, I am the maintainer of oneChannelGUI, a graphical inteface to various Bioconductor packages. I am putting some effort in increasing the capability of oneChannelGUI in the analysis of exon-arrays. Recently Xing et al . have published a nice paper on a new tool for exon arrays analysis (MADS: a new and improved method for analysis of differential alternative splicing by exon-tiling microarrays. RNA. 2008 Aug;14(8):1470-9). This tool was subsequently embedded in the JETTA software http://gluegrant1.stanford.edu/~junhee/JETTA/ , which is also available as R package for win/mac/linux platforms. I am working on interfacing this libray in oneChannelGUI, however I need some advice regarding the insertion in Biocondutor of libraries that are not part of it. Since in one of my function I call jetta library, I need to declare it in the suggested packages in the DESCRIPTION file of oneChannelGUI. In my function I have implemented a routine that download and install from the http://gluegrant1.stanford.edu/~junhee/JETTA/ web page the specific platform package. Therefore the users interested to use jetta could handle it. However, if I will load the package on the devel repository of Bioconductor, the presence in the suggested packages of jetta will produce a warning. How can I handle this problem? Is this acceptable by Bioconductor? Is there any way to avoid the warning? e.g. ask to the Biocore team to install the jetta lib on the repository. Is this feasible? I am open to any suggestion Cheers Raffaele