R package¶

The StreamFind R package provides the existing R interface for DuckDB-backed mass-spectrometry and non-target-screening workflows, including the Shiny application.

Separate interface

The R package is separate from the native C++ and Rust packages. Use this page for R installation and usage; use Releases for the native packages and MCP servers.

Installation¶

The package requires R and a suitable native build toolchain. On Windows this generally includes RTools.

Install from GitHub:

options(timeout = 600)
if (!requireNamespace("remotes", quietly = TRUE)) {
  install.packages("remotes")
}
remotes::install_github("ricardo-cunha/streamfind", subdir = "bindings/r")

The companion example-data package can be installed separately:

remotes::install_github("odea-project/streamfind.data")

Project workflow¶

Open or create a non-target-analysis project, define an ordered workflow, and run it:

library(streamfind)

nta <- open_ProjectNonTargetAnalysis(
  db = "streamfind.duckdb",
  project_id = "nta_demo"
)

workflow <- Workflow(list(
  Method_NonTargetAnalysis_FindFeatures(),
  Method_NonTargetAnalysis_LoadFeaturesMS1(),
  Method_NonTargetAnalysis_LoadFeaturesMS2(),
  Method_NonTargetAnalysis_GroupFeatures(),
  Method_NonTargetAnalysis_FillFeatures(),
  Method_NonTargetAnalysis_FilterFeatures()
))

set_workflow(nta, workflow)
run_workflow(nta)

features <- get_features(nta)
plot_features_count(nta)

Workflow methods must be ordered according to their prerequisites. The package help pages document each method's parameters and expected project state.

Chromatograms¶

Chromatogram workflows use the corresponding method constructors:

load_chromatograms <- Method_MassSpecChromatograms_LoadChromatograms()
filter_chromatograms <-
  Method_MassSpecChromatograms_FilterChromatogramsRetentionTime(
    rtmin = 0,
    rtmax = 600
  )

workflow <- Workflow(list(load_chromatograms, filter_chromatograms))
set_workflow(nta, workflow)
run_workflow(nta)

Shiny application¶

The package includes a Shiny application for browsing projects, analyses, workflows, chromatograms, spectra, and non-target-analysis results:

library(streamfind)

nta <- open_ProjectNonTargetAnalysis(
  db = "streamfind.duckdb",
  project_id = "nta_demo"
)
nta$run_app()

The application opens in a browser when run locally.

Container image¶

A container image for the R package is available from Docker Hub. Consult the image documentation for the current image configuration and runtime options.