Article-Journal | Li Rui - Shanghaitech University

TRACE: An Integrated Isotope-Tracing Framework for Metabolite Validation and Nutrient Fate Mapping

Tue, 18 Aug 2026 00:00:00 +0000

TRACE is an integrated experimental and computational framework for untargeted LC–MS metabolomics. Fully labeled ¹³C-glucose / ¹⁵N-ammonium cultures supply high-confidence CN bundles that seed a peak–peak network; dynamic m/z and retention-time tolerances are learned from those seeds, then unlabeled nutrients are mapped by isotopic dilution.

The accompanying R package is TRACE (GitHub: DrRuiLi/TRACE). A project-level overview, including workflow figures, is on the TRACE project page.

Why a new framework

Untargeted LC–MS surveys microbial metabolism at scale, but two problems remain coupled: many detected peaks are adducts, fragments, or contaminants rather than cell-made metabolites, and classical tracing follows one labeled substrate at a time. TRACE treats isotope-confirmed peaks as network seeds, calibrates error tolerances from those seeds, and asks how cells partition competing carbon and nitrogen sources.

Main results

Calibrated networks. Complete unlabeled / ¹⁵N / ¹³C / dual-label CN bundles (Pearson ρ > 0.75) collapsed ~25,000 features to 1,098 high-confidence seeds in a representative positive-mode data set, from which data-driven m/z and RT tolerances are learned.
Better assignment than pairwise methods. Global network assignment grouped adducts, fragments, and isotopologues of one metabolite, removed large m/z/RT errors seen in PAVE, and recovered missed features. About 70% of TRACE-annotated peaks matched HMDB, YMDB, KEGG, or an in-house database.
Sensitivity is not biology. From Q Exactive Plus to Excedion Pro, total features rose 331% while TRACE-validated metabolites rose 117%, and the share of high-confidence annotations fell from 2.94% to 1.48%.
Nutrient fate under competition. Unlabeled leucine, threonine, tryptophan, adenine, uracil, acetate, or a 14-compound mixture diluted the labeled background in a pathway-selective way. Glutathione tracked linear amino-acid assembly; NAD⁺ revealed concurrent salvage, de novo, and independent nitrogen entry.

Software versions

Version	Role
`v1.0.0`	Code used for this article
`main` (1.1.0+)	Ongoing development

devtools::install_github("DrRuiLi/TRACE", ref = "v1.0.0")

Metabolomics and Machine Learning Identify Metabolic Differences and Potential Biomarkers for Frequent versus Infrequent Gout Flares

Wed, 01 Nov 2023 00:00:00 +0000

Click the Cite button above to demo the feature to enable visitors to import publication metadata into their reference management software.

Create your slides in Markdown - click the Slides button to check out the example.

Add the publication’s full text or supplementary notes here. You can use rich formatting such as including code, math, and images.

Metabolomics in Hyperuricemia and Gout

Tue, 21 Mar 2023 00:00:00 +0000

Click the Cite button above to demo the feature to enable visitors to import publication metadata into their reference management software.

Create your slides in Markdown - click the Slides button to check out the example.

Add the publication’s full text or supplementary notes here. You can use rich formatting such as including code, math, and images.

Plasma Metabolomics Reveals Systemic Metabolic Alterations of Subclinical and Clinical Hypothyroidism

Sun, 01 Jan 2023 00:00:00 +0000

Click the Cite button above to demo the feature to enable visitors to import publication metadata into their reference management software.

Create your slides in Markdown - click the Slides button to check out the example.

Add the publication’s full text or supplementary notes here. You can use rich formatting such as including code, math, and images.