New software paper: GenomicInteractions

Screen Shot 2016-05-03 at 16.38.11
The paper associated with our GenomicInteractions R package is now out! You can read it here on the BMC Genomics website.

The GenomicInteractions package came about because several members of the group (me, Nathan, Malcolm and Anja) were working with genomic interaction data, and realised that there were few tools designed to handle this kind of data in R. In particular, we wanted a package that could handle both ChIA-PET data (which Nathan and Malcolm were using) and HiC data (which I was using). Nathan designed the initial data structure and functions, and we all contributed to package development via GitHub. Github was incredibly useful for this project – it would have been basically impossible to collaborate on code in the same way without it.

Writing the package was a great learning experience – we learnt how to make an R package, of course, but also learnt more about tools available for R development (e.g. devtools, roxygen), about collaborating using git and Github, and about how to write documentation. I also took the opportunity to learn how to write unit tests using the testthat package (although the test coverage still isn’t as high as I’d like).

The package has been available in Bioconductor since October 2014, but since that first version we’ve changed the data structure to be more similar to other Bioconductor objects, and added a function to plot interaction data with Gviz. These are described in the paper. We’re still working on improving the package and adding more functionality, so there’ll probably be more updates here in the future!