Entering edit mode
I help manage all of the R package installations for my team's offline systems. Is there a method to determine which packages or dependencies require an internet connection to function without manually testing the package on an offline system?
For example, I have noticed that there are packages like biomaRt that require an internet connection to function.
Many packages make some use of the internet but do not require it for all purposes. For example, the limma package can be 99% used offline but limma::kegga() normally requires an internet connection to access KEGG pathway definitions. Even kegga() can be used offline if the KEGG pathways are already downloaded.
Exactly, that would be one of our our "alternative installation" plans for a package requiring internet! Thank you!! Do you happen to know how to quickly identify which packages have a function(s) that will require an internet connection? I checked the Bioconductor pages for packages like biomaRt but it was not defined until I looked in the manual
No, I do not know any way to determine that.
Internet usage generally follows analysis stage rather than packages used. If you are downloading data or annotation or installing software, then you need the internet. If you are purely analysing data, then you don't need the internet.
Installing a version of Bioconductor that can be relied on to not require the internet at any stage -- that hardly seems possible.