Skill v1.0.0
currentLLM-judged scan95/100version: "1.0.0" name: bioconductor-ompbam description: This packages provides C++ header files for developers wishing to create R packages that processes BAM files. ompBAM automates file access, memory management, and handling of multiple threads 'behind the scenes', so developers can focus on creating domain-specific functionality. The included vignette contains detailed documentation of this API, including quick-start instructions to create a new ompBAM-based package, and step-by-step explanation of the functionality behind the example packaged in when_to_use: Use when: This packages provides C++ header files for developers wishing to create R packages that processes BAM files; ompBAM API Documentation; Single-cell RNA-seq analysis (ompBAM). Not for: Requires R ≥ 4.0.0 and Bioconductor ≥ 3.16 user-invocable: false
ompBAM
Workflows
Standard Workflow
library(ompBAM)# Create a new package template that compiles with ompBAMpkg_path <- file.path(tempdir(), "MyPkg")use_ompBAM(pkg_path)# Retrieve the path to an example BAM file provided by the packagebam_path <- ompBAM::example_BAM("Unsorted")
Input: A destination directory path for the new package. Output: A fully configured R package directory structure containing C++ templates and configuration files linked to ompBAM.
When to Use
- Developing R packages that require high-performance, multi-threaded processing of BAM files using C++ and OpenMP.
- Writing custom BAM parsing algorithms where you want
ompBAMto handle file access, BGZF decompression, and thread management. - Creating packages that need to run on unsorted and unindexed BAM files (e.g., using
pbam_inandpbam1_tAPIs).
When NOT to Use
- For standard BAM manipulation or filtering in R without writing C++ code; use
RsamtoolsorGenomicAlignmentsinstead. - For querying small genomic regions of sorted and indexed BAM files; use
htslibdirectly asompBAMis optimized for whole-file sequential reads.
Data Requirements
- Input BAM files (can be unsorted and unindexed).
- A package development environment with
Rcppand a C++ compiler supporting OpenMP (e.g.,gccorclangwith OpenMP libraries).
Key Parameters
- path (in
use_ompBAM): The directory path where the new package project will be created. - bam_file: Path to the BAM file to be processed.
- n_threads: Number of threads to use for parallel execution.
Best Practices
- Run
devtools::document()on the newly created package to generate theNAMESPACEand export the R wrapper functions. - Use the
#ifdef _OPENMPdirective in C++ code to ensure compatibility with systems that do not have OpenMP installed (such as default macOS configurations). - Call
pbam1_t::realize()if you need to store reads in memory across multiple calls ofpbam_in::fillReads(), as virtual reads point to temporary buffers.
Common Pitfalls
- macOS Compilation Failure: macOS does not support OpenMP natively. Fix: Install
libompvia Homebrew (brew install libomp) and configure compiler flags. - Memory Corruption with Virtual Reads: Accessing
pbam1_tobjects after callingfillReads()leads to segmentation faults. Fix: Callrealize()on thepbam1_tobject to copy the data into persistent memory. - Requesting Too Many Threads: Requesting more threads than available can degrade performance. Fix: Use a helper function to cap requested threads at
omp_get_max_threads().
Alternatives
- Rsamtools: For high-level BAM file manipulation and querying in R.
- GenomicAlignments: For representing and manipulating genomic alignments in R.
- Rcpp: For general C++ integration in R packages without built-in BAM parsing support.
Citations
- Alex C H Wong (2026). ompBAM API Documentation.
References
- Homepage: https://bioconductor.org/packages/ompbam
- Vignette: https://bioconductor.org/packages/release/bioc/vignettes/ompbam/inst/doc/ompBAM-API-Docs.html