Parallelize 'GenomicAlignments' functions

Henrik Bengtsson

+ =

The futurize package allows you to easily turn sequential code into parallel code by piping the sequential code to the futurize() function. Easy!

TL;DR

library(futurize)
plan(multisession)
library(GenomicAlignments)

se <- summarizeOverlaps(features, bam_files) |> futurize()

Introduction

This vignette demonstrates how to use this approach to parallelize the GenomicAlignments functions.

The GenomicAlignments Bioconductor package provides efficient representation and manipulation of short genomic alignments. The summarizeOverlaps() function counts the number of reads that map to each feature (e.g. gene or exon) from one or more BAM files. When called with a BamFileList, the work is distributed across BAM files using bplapply(), which can be parallelized.

Example: Running summarizeOverlaps() in parallel

The summarizeOverlaps() function counts reads overlapping genomic features across multiple BAM files:

library(GenomicAlignments)
library(Rsamtools)

bam_files <- BamFileList(c("sample1.bam", "sample2.bam", "sample3.bam"))
features <- GRanges("chr1",
  IRanges(start = c(1, 1000, 2000), end = c(500, 1500, 2500))
)

se <- summarizeOverlaps(features, bam_files)

Here summarizeOverlaps() processes BAM files sequentially, but we can easily make it process them in parallel by piping to futurize():

library(futurize)

se <- summarizeOverlaps(features, bam_files) |> futurize()

This will distribute the BAM file processing across the available parallel workers, given that we have set up parallel workers, e.g.

plan(multisession)

The built-in multisession backend parallelizes on your local computer and works on all operating systems. There are other parallel backends to choose from, including alternatives to parallelize locally as well as distributed across remote machines, e.g.

plan(future.mirai::mirai_multisession)

and

plan(future.batchtools::batchtools_slurm)

Supported Functions

The following GenomicAlignments functions are supported by futurize():

• summarizeOverlaps()