Bioconductor and R compatibility
There is a table of Bioconductor and R compatibility here
but it is flawed in an easily fixable way. It doesn't list the versions beyond the first two numbers.
And so I installed R version 3.3.2, thinking it would work with the latest Bioconductor release v 3.4 but it doesn't as evidenced by
DESeq2, which is "a Bioconductor package"(Source:here) not being able to be installed. When I look on the other pages , I see (AT THE TIME ORIGINALLY WRITTEN AROUND APRIL 2017)
"Bioconductor is version 3.4; it works with R version 3.3.1."
Additional Dependencies OFTEN REQUIRED
In the process of Dockerizing, I noted many needed dependencies that I only learned about by forensics on install messages and combing the web that were needed for installation of R itself and then for proper installation of Bioconductory packages, such as
DESeq. This goes just beyond trying to get this working in Docker containers. Indeed, many others have stumbled on these uncommon (perhaps?), in-development depenencies as evidenced by here, here, here, here, and here, and here whose need is not clearly documented IMO in the ‘R Installation and Administration’ manual or the Installation instructions at Bioconductor: Home: Install. For example, cyrptic notes here about
PCRE suggest consulting the ‘R Installation and Administration’ manual and talks about building
PCRE, but no where at Bioconductor: Home: Install is it mentioned or suggested to use a package manager to install
As of early 2017, these were some of such dependencies I noted concerned
XML2, and required the following:
In fact I would say highly recommend just installing these before you begin the process of trying to install R to save youself many headaches.
See the 'Install R/Install R Packages 'section of this Dockerfile if you need additional guidance in how to install these. Leave off the leading, Docker-specfic
RUN when you run these commands on the command line.
Resources concerning Bioconductor and R compatibility
The most recent one (AT THE TIME ORIGINALLY WRITTEN AROUND APRIL 2017) is 3.4, but the big issue I note is at the top it says
"Bioconductor 3.4 Release Schedule This release will use R-3.3.1 (Bug in Your Hair) unless 3.3.2 comes out in time (would need to be at least 1 week before release day though)."
Plus going to here from that page above, I see
"The current release of Bioconductor is version 3.4; it works with R version 3.3.1. Users of older R and Bioconductor users must update their installation to take advantage of new features and to access packages that have been added to Bioconductor since the last release."
I see on the side the have a link to Release announcements and there is a table there if compatibility with preface
"Each Bioconductor release is designed to work with a specific version of R. The following table summarizes the relationship, and links to packages designed to work with the corresponding R / Bioconductor version."
But that table is flawed in my view. See above.
Tracking packages in relation to R and Bioconductor versions:
You can easily view currently installed packages and versions three lines of code found here:
ip <- as.data.frame(installed.packages()[,c(1,3:4)]) rownames(ip) <- NULL ip <- ip[is.na(ip$Priority),1:2,drop=FALSE] print(ip, row.names=FALSE)
However... Keeping track of the individual packages I am using sometimes comes as an after thought, but at least I usually have to track what versions of R and Bioconductor I am using in order to set up the right RStudio server instance when I cloud compute and try to remember to at least record that. Luckily with that in hand, you can easily get that information... Searching
check which version of a package for version of bioconductorlead me to https://www.biostars.org/p/121699/#121721 From Martin Morgan's answer at lead to http://bioconductor.org/help/ from the link in text as "See the 'Previous versions' box on the help overview page to find the version of a package associated with a particular version of R". From
Previous Releasesbox you can click on the versions you had previously used (HOPEFULLY NOTED) and actually check exactly what versions of specific packages were involved. For example, at the time of writing this the current version of Bioconductor is 3.6 to work with R version 3.4.2 (see here) and
tximportcurrent version comes up as 1.6.0 if I just google
bioconductor tximportand go to the tximport page, but if I want to see what version I had used before when I had worked with Bioconductor version 3.4; R version 3.3.1, I can select "3.4 (3.3)" from the "Previous Releases" box at http://bioconductor.org/help/ and be lead to http://bioconductor.org/packages/3.4/BiocViews.html#___Software where all the
Packages found under Softwarewill load up, and I can click on specific ones and see the versions that were associated. For the tximport example if, I click on
[tximport](http://bioconductor.org/packages/3.4/bioc/html/tximport.html)and be taken to http://bioconductor.org/packages/3.4/bioc/html/tximport.html where I'll see the version at that time was 1.2.0.