Answers are grounded in this paper and cite their sources.
Introduction Untargeted liquid chromatography–mass spectrometry (LC–MS) metabolomics and lipidomics generate thousands of features per sample, many of which fall outside the instrument’s linear dynamic range or reflect injection artefacts. Despite international quality-control (QC) guidelines, per-feature QC filtering between feature extraction and statistical analysis remains poorly standardized. Objectives We present Aletheia, an open-source pipeline for QC dilution-based feature filtering, describe its behaviour on a reproducible synthetic plasma dataset, and check its compatibility with the notame workflow. Methods A synthetic plasma reference was analysed using a five-level QC dilution series in triplicate in both polarities, alongside pooled QC injections. Aletheia combines linearity assessment, replicate diagnostics, non-monotonic response flagging, and QC-pool repeatability, and imports MetaboScape, MZmine 3, XCMS and MS-DIAL tables. Results Of 11,694 features that had passed a pooled-QC filter (coefficient of variation of 20% or less), 10,732 (91.8%) were retained, including 1,174 accepted by a two-point positive-slope rule that does not test linearity and 260 (2.2%) flagged as non-monotonic, whereas 962 (8.2%) showed no dilution response and were rejected. Combining linearity and repeatability criteria, 8,173 features (69.9%) passed both. Conclusion Aletheia offers a transparent and reproducible implementation of QC dilution-based feature filtering, with feature-level labels and machine-readable audit trails. It is open source under the MIT license and available at https://github.com/aperalbom/Aletheia .
No takes yet. Share an insight, caveat, or question.
Peralbo‐Molina et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: