rtracklayer | R interface to genome annotation files | Genomics library

Thankfully I was able to fix the issue I was having with the installation and loading of the Bioconductor GenomicFeatures from within python3.8 via rpy2 .I spent hours reaching that solution so I wanted to share the solution of what I did so that if whom end up with the same issue can find it.

Basically, when tracking the error log, it seems that the actual error was pointing to this particular error line:

Went into .bash_profile and deleted the R_LIBS_USER variable. It's now working. Suggested by Johannes Koester.

With ranges being different lengths, I'm assuming you want these randomly chosen positions to be proportional to the length of the segment. In other words, selection is uniform based on actual base pairs within the ranges. Otherwise you'll be over-representing small ranges (higher marker density) and under-representing large ranges (lower marker density).

Here's a data.table solution that can do one thousand sites pretty much instantly, and one million random sites in about 10 seconds on my machine. It randomly samples the number of sites you want, first by sampling rows (weighted by each row's range size), then sampling uniformly within that range.

One can get help for the biocLite function with:

I'm basically on the same system as you. I eventually got rtracklayer to load, but it required perseverance and reading a lot of error messages and waiting. The hangup was in getting Rsamtools to load and for that I removed old library copies, install XCode Command Line Tools (because of error with XML) and reloaded source("https://bioconductor.org/biocLite.R") after restarting. I also installed clang4 as per: https://cran.r-project.org/bin/macosx/tools/. Not sure which of these were the solution, and it's possible that several of them were needed.