MetaDilutionR: An R Package for Data-Driven Determination of Optimal Plasma Dilution in Untargeted Metabolomics to Achieve Maximum Metabolite Coverage with ESI Linearity

High-resolution mass spectrometry (HRMS) instruments for untargeted metabolomics typically offer a linear dynamic range spanning approximately 4 orders of magnitude. However, biological samples contain metabolites spanning concentration ranges far exceeding this window, making dilution optimization critical for reliable quantification. Despite its importance, dilution selection in untargeted workflows is rarely standardized and is often determined empirically through manual inspection, leading to operation outside the linear dynamic range, ion suppression, and compromised reproducibility. To address this, we present MetaDilutionR, an open-source R package that standardizes dilution optimization by systematically evaluating electrospray ionization (ESI) linearity using plasma as a model. MetaDilutionR automates dilution assessments, applying user-adjustable slope and R2 thresholds to classify features as linear or nonlinear and executes the complete analysis via a single function call, ensuring algorithmic reproducibility across users and platforms. The package generates comprehensive outputs, including log2-transformed data, a summary of linear features with their optimal dilution ranges, nonlinear features highlighting potential ion suppression or detector saturation, detailed evaluations across dilution scenarios, and visual regression plot reports. Benchmarking against three established metabolomics workflows demonstrated that the R2-slope criterion of MetaDilutionR reduces false-positive linear assignments. Cross-platform applicability of the algorithm on an independent GC-MS data set confirmed consistent classification performance beyond LC-HRMS. By facilitating systematic identification of metabolite-specific optimal dilution conditions, MetaDilutionR enables metabolites to be quantified within their linear dynamic range─a prerequisite for reliable quantification─thereby enhancing reproducibility and consistency of downstream validation, making it readily integrable into existing metabolomics workflows.