Unable to Install minfi with BioConductor on R Studio Cloud

[jg2335]

I have tried multiple different versions and get different errors every time. For example, when try to install minfi on 4.1.3 I get the following error.

installing to /cloud/lib/x86_64-pc-linux-gnu-library/4.1/00LOCK-genefilter/00new/genefilter/libs
** R
** data
*** moving datasets to lazyload DB
** inst
** byte-compile and prepare package for lazy loading
** help
*** installing help indices
** building package indices
** installing vignettes
** testing if installed package can be loaded from temporary location
** checking absolute paths in shared objects and dynamic libraries
** testing if installed package can be loaded from final location
** testing if installed package keeps a record of temporary installation path
* DONE (genefilter)
* installing *source* package ‘GenomicFeatures’ ...
** using staged installation
** R
** inst
** byte-compile and prepare package for lazy loading
Killed
ERROR: lazy loading failed for package ‘GenomicFeatures’
* removing ‘/cloud/lib/x86_64-pc-linux-gnu-library/4.1/GenomicFeatures’
ERROR: dependency ‘GenomicFeatures’ is not available for package ‘bumphunter’
* removing ‘/cloud/lib/x86_64-pc-linux-gnu-library/4.1/bumphunter’
ERROR: dependency ‘bumphunter’ is not available for package ‘minfi’
* removing ‘/cloud/lib/x86_64-pc-linux-gnu-library/4.1/minfi’

The downloaded source packages are in
	‘/tmp/RtmpgZwi0s/downloaded_packages’
Installation paths not writeable, unable to update packages
  path: /opt/R/4.1.3/lib/R/library
  packages:
    cluster, MASS, Matrix, mgcv, nlme, survival
Warning messages:
1: In install.packages(...) :
  installation of package ‘GenomicFeatures’ had non-zero exit status
2: In install.packages(...) :
  installation of package ‘bumphunter’ had non-zero exit status
3: In install.packages(...) :
  installation of package ‘minfi’ had non-zero exit status

When I try on 4.2.0 I get this error

installing to /cloud/lib/x86_64-pc-linux-gnu-library/4.2/00LOCK-rtracklayer/00new/rtracklayer/libs
** R
** data
** demo
** inst
** byte-compile and prepare package for lazy loading
Killed
ERROR: lazy loading failed for package ‘rtracklayer’
* removing ‘/cloud/lib/x86_64-pc-linux-gnu-library/4.2/rtracklayer’

The downloaded source packages are in
	‘/tmp/RtmpCru8my/downloaded_packages’
Installation paths not writeable, unable to update packages
  path: /opt/R/4.2.0/lib/R/library
  packages:
    MASS, nlme, survival
Warning message:
In install.packages(...) :
  installation of package ‘rtracklayer’ had non-zero exit status

I am using the free tier of R Studio cloud but am not seeing any RAM issues. I literally start a new workspace/project. And run the following code:

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install(version = "3.15") # or 3.13 for R version 4.1

BiocManager::install("minfi")

And I just cant get it to work. Does anyone know what is going on? Thank you!

[AlexisW]

@jg2335

Killed
ERROR: lazy loading failed for package ‘GenomicFeatures’

I'm pretty sure this is what happens when you run out of memory. And running the same code on an RStudio Cloud instance, it does look like the RAM keeps coming very close to the limit. I don't have a paid instance on hand to check what happens with more than 1GB RAM.

[andresrcs]

@jg2335

** byte-compile and prepare package for lazy loading
Killed

You are running out of RAM memory while compiling, to overcome package installation issues derived from memory restrictions on the free tier, RStudio Cloud uses the public RStudio Package Manager to host precompiled binaries but as far as I know they only include CRAN packages, not bio Bioconductor.

To overcome this you would need a paid account to increase the memory limit for your project.

[andresrcs]

I take it back, RSPM does include Bioconductor binaries, and you are not running out of RAM memory, there is a different issue while installing rtracklayer even from the RSPM binary, I suspect this is because the package was built for R 4.2, and not 4.2.1 and it is not compatible with the latest stats version but I'm not sure and I haven't found a workaround. sorry!