If we're going to parallelize this, it is better to use BiocParallel so that users can either run sequentially, in parallel with multiworkers, in parallel with multiprocess, or some other supported backend:
Separately, I had implemented a toy version just using lapply() for vectors of files, which I can PR separately. The idea was for the functions to have a bin/summarize argument that could emit a GRanges or DataFrame suitable for arbitrary model fits, per Samantha's Q.
If we're going to parallelize this, it is better to use BiocParallel so that users can either run sequentially, in parallel with multiworkers, in parallel with multiprocess, or some other supported backend:
http://lcolladotor.github.io/2016/03/07/biocparallel/
Separately, I had implemented a toy version just using lapply() for vectors of files, which I can PR separately. The idea was for the functions to have a bin/summarize argument that could emit a GRanges or DataFrame suitable for arbitrary model fits, per Samantha's Q.