R Package API Usage

R Package API Usage

Summary

Programmatically invoke R package functions to perform metabolomics quality control analyses and generate standardized reports. This skill encompasses reading data, building experiment objects, executing analysis pipelines, and exporting results through mzQuality's documented API.

When to use

When you have metabolomics data (tab-delimited text or SummarizedExperiment object) and need to apply batch correction, outlier detection, internal standard recommendation, and quality filtering at scale or in non-interactive workflows. Use this skill when makeSummaryReport and makeCompoundReport flags must be set programmatically to control report output types.

When NOT to use

• Input data lacks mandatory columns (compound ID, aliquot/sample name, assay area values, sample type) or cannot be coerced to SummarizedExperiment format.
• Analysis requires interactive exploration or real-time parameter tuning; use mzQualityDashboard Shiny application instead.
• No Quality Control (QC) samples are present in the dataset; Rosner outlier test and batch correction rely on pooled QC replicates.

Inputs

• Tab-delimited text file (Sciex OS export or user-formatted with compound, aliquot, assay, type columns)
• SummarizedExperiment object (from existing R-based pipeline, e.g., xcms)
• Pre-analyzed experiment object (result of doAnalysis)

Outputs

• SummarizedExperiment object with added assays (ratio, ratio_corrected, background %, matrix effect, RSD_QC, median area, presence)
• rowData slot with use (TRUE/FALSE) column and compound reliability flags (High Confidence, Caution, Low SNR)
• colData slot with use column and sample outlier status
• Project folder containing Plots/ (visualization PNGs/PDFs), Reports/ (HTML summary and compound reports, tab-delimited matrices), and Excel workbook

How to apply

Load metabolomics data using readData() to validate column structure (mandatory: compound, aliquot, assay, type columns); construct a SummarizedExperiment via buildExperiment(); invoke doAnalysis() with threshold parameters (RSDQC, background signal %, QC presence) to compute batch-corrected ratios, detect outliers via Rosner Test on QC samples, and flag unreliable compounds; subset using rowData(exp)$use and exp$use to retain high-confidence results; call createReports() with makeSummaryReport=TRUE and makeCompoundReport=TRUE to export a Project folder with Plots/ and Reports/ subdirectories containing HTML visualizations, tab-delimited matrices (assays, rowData, colData), and Excel workbook.

Related tools

• mzQuality (Primary R package providing readData(), buildExperiment(), doAnalysis(), and createReports() API functions for quality control analysis) — https://github.com/hankemeierlab/mzQuality
• SummarizedExperiment (Bioconductor class for storing assay data, row metadata (compounds), and column metadata (samples/aliquots) used internally by mzQuality) — https://bioconductor.org/packages/release/bioc/html/SummarizedExperiment.html
• xcms (R-based metabolomics preprocessing pipeline that outputs SummarizedExperiment objects compatible with mzQuality input)
• mzQualityDashboard (Interactive Shiny application wrapper around mzQuality API for users without R programming experience) — https://github.com/hankemeierlab/mzQualityDashboard

Examples

path <- system.file("extdata", "example.tsv", package = "mzQuality"); exp <- doAnalysis(buildExperiment(readData(path))); exp <- exp[rowData(exp)$use, exp$use]; createReports(exp, makeSummaryReport=TRUE, makeCompoundReport=TRUE, dir="./analysis_output")

Evaluation signals

• SummarizedExperiment object contains new assay matrices (ratio, ratio_corrected, background_percent, matrix_effect, rsd_qc, median_area, presence) matching sample count and compound count
• rowData includes use column with all TRUE or FALSE values; compounds marked FALSE have RSDQC > threshold, background % > threshold, or < presence threshold
• colData includes use column; QC samples marked FALSE are identified as statistical outliers by Rosner Test on compound/IS ratios
• Project folder exists with Plots/ and Reports/ subdirectories; Reports/ contains summary and compound HTML files; tab-delimited files (assays.tsv, rowData.tsv, colData.tsv) are readable and match SummarizedExperiment dimensions
• Excel workbook generated contains sheets for each assay, rowData, and colData with no missing values in key columns (compound ID, sample ID, ratios, flags)

Limitations

• Requires pooled Quality Control (QC) samples; analysis will not perform outlier detection or reliable batch correction without QC replicates.
• Tab-delimited input must have exact column names and format; readData() validation is strict and will reject malformed files.
• Internal Standard assignment must be pre-specified; mzQuality recommends standards but does not automate selection if multiple candidates exist.
• Batch correction assumes systematic variation; if batch effect is sample-type-dependent or non-linear, results may be suboptimal.
• No changelog provided in repository; API stability and parameter names across versions are not documented.

Evidence

• [readme] buildExperiment and doAnalysis workflow: "Once your files are ready, you can use the readData function to read in your data. It will check if all mandatory columns are present and if the data is in the correct format. Finally, the"
• [readme] doAnalysis steps and outputs: "The doAnalysis function will perform the following steps: 1. Calculate the ratio between the compounds and assigned internal standards, 2. Perform batch correction using the pooled study quality"
• [other] createReports output structure: "createReports generates a Project folder organized into Plots and Reports subdirectories containing analysis results, including summary reports, compound reports, and tab-delimited text files for"
• [readme] SummarizedExperiment selection workflow: "mzQuality adds a column called use in both the rowData and colData slots of the SummarizedExperiment. These contain either a TRUE or FALSE value, indicating if the compound or sample is"
• [readme] Outlier detection and sample filtering: "It bases the decision for samples on the outcome of the Rosner Test, which tests for statistical outliers in QC samples."
• [intro] Input data format requirement: "it features import of data from a variety of formats, including a generalized tab-delimited format and Sciex OS text exports"