Summary
A bioRxiv preprint presents Rsearch, an R package that integrates VSEARCH with visualisation, parameter optimisation and conversion tools for metabarcoding analysis. In a mock-community comparison with DADA2, the two pipelines produced abundance profiles that closely matched the expected composition, with differences in OTU counts and computational performance.
Researchers have presented Rsearch, an R package that brings the core functionality of the VSEARCH sequence-analysis toolkit into the R environment. The package adds visualisation, parameter optimisation and conversion functions intended to connect metabarcoding workflows with statistical and ecological analysis in R.
The work is described in a bioRxiv preprint by researchers from the Norwegian University of Life Sciences and the University of Oslo. Rsearch is available through the Comprehensive R Archive Network (CRAN), while its development version is hosted on GitHub.
What Rsearch adds to metabarcoding workflows
Metabarcoding uses DNA sequences recovered from an environmental or mixed biological sample to identify the organisms present. A key stage is grouping or denoising similar sequences and summarising them as operational taxonomic units, or OTUs. The resulting abundance table can then be examined with statistical and ecological software.
VSEARCH is used for several of these sequence-processing tasks. Rsearch makes VSEARCH functions accessible from within R rather than requiring users to move between separate command-line and statistical environments. The package also includes tools for selecting trimming and filtering parameters systematically. These settings determine which parts of sequencing reads are retained and which reads are discarded before downstream analysis.
The authors also provide conversion functions for compatibility with other R packages, including phyloseq. This is intended to make it easier to carry processed sequence data into later ecological and statistical analyses.
Results from the DADA2 comparison
The study compared Rsearch with DADA2 using mock-community data. A mock community contains a defined mixture of organisms or sequences, giving researchers an expected composition against which an analysis pipeline can be assessed.
Both pipelines produced relative-abundance profiles that were highly correlated with the expected composition. The methods nevertheless differed in how they represented diversity. Rsearch identified fewer OTUs, and those OTUs were more consistently prevalent across samples. In the authors’ analysis, DADA2 appeared to estimate a higher level of diversity by splitting sequences into a larger number of OTUs.
The computational results also varied by function. The vs_cluster_unoise method implemented in Rsearch was the fastest among the clustering and denoising methods evaluated. Other Rsearch functions had runtimes comparable to DADA2.
These findings make Rsearch relevant not because it replaces every part of an existing workflow, but because it combines VSEARCH-based processing with R’s analysis ecosystem. The authors describe parameter optimisation as particularly useful for researchers who may not otherwise have a clear strategy for choosing trimming and filtering settings.
The evidence is currently from a bioRxiv preprint and includes a mock-community comparison. The reported performance and OTU differences should therefore be interpreted in the context of that test setting, while broader use across metabarcoding datasets will provide additional evidence about how the package behaves in different workflows.