MetaboLightsapplication/xmlftp://ftp.ebi.ac.uk/pub/databases/metabolights/studies/public/MTBLS12655/m_MTBLS12655_LC-MS___metabolite_profiling_v2_maf.tsvftp://ftp.ebi.ac.uk/pub/databases/metabolights/studies/public/MTBLS12655/s_MTBLS12655.txtftp://ftp.ebi.ac.uk/pub/databases/metabolights/studies/public/MTBLS12655/a_MTBLS12655_LC-MS___metabolite_profiling.txtftp://ftp.ebi.ac.uk/pub/databases/metabolights/studies/public/MTBLS12655/i_Investigation.txtprimaryOK200ftp://ftp.ebi.ac.uk/pub/databases/metabolights/studies/public/MTBLS12655<p>Metabolite identification and quantification were conducted using known reference standards and predefined MRM transitions from Waters' internal database within MassLynx. Identified metabolites were cross-referenced with public databases such as HMDB and KEGG for confirmation of identity and pathway mapping.</p>MetaboLightsPublicLiquid Chromatography MS -<p>Chromatographic separation was performed using the ACQUITY UPLC system (Waters, USA). The Scherzo SM-C18 column (2 x 100 mm, 3 µm; Imtakt, Japan) was used with a flow rate of 0.2 mL/min. The mobile phase consisted of 0.1% formic acid in water (A) and acetonitrile (B). The gradient started at 5% B, ramped to 100% over 5 min, held for 3 min and re-equilibrated at 5% B. Injection volume was 1 µL.</p>Identification and Validation of Plasma Lipid Biomarkers for Oral Cancer Diagnosis.National cancer centerYeon-hee Kimblood plasmamass spectrometry assay<p>For targeted polar metabolomics, 20 µL of plasma was extracted with 150 µL of acetonitrile and 30 µL of water. The mixture was centrifuged, and the supernatant was diluted using acetonitrile:water (75:25, v/v) containing internal standards.</p>Homo sapienshttps://www.ebi.ac.uk/metabolights/MTBLS12655Kim Yeon-Hee. National cancer center. 323, Ilsan-ro, Ilsandong-gu, Goyang-si, Gyeonggi-do. yh0227@ncc.re.kr. 82319202215.Kim Mi Kyung. National cancer center. 323, Ilsan-ro, Ilsandong-gu, Goyang-si, Gyeonggi-do. alrud@ncc.re.kr. 82329202202.<p>Raw mass spectrometry data were processed using Waters MassLynx (version 4.2). Data preprocessing included peak integration, alignment and quantification. Metabolite intensities were normalized and exported as quantitative tables. Missing values were imputed using a limit of detection approach (1/5 of minimum detected intensity). Statistical analysis, normalization, log transformation and Pareto scaling were performed using MetaboAnalyst 5.0 (https://www.metaboanalyst.ca).</p>Canceryh0227@ncc.re.kr<p>Plasma samples were collected from 650 human participants, including patients with newly diagnosed oral cancer and age and sexmatched healthy controls. Blood was drawn into EDTA tubes, immediately frozen, and centrifuged at 845 x g for 20 min at 4 °C to isolate plasma. Samples were stored at -80 °C until analysis. All procedures were IRB approved (NCC2016-0147) and informed consent was obtained from all participants.</p>Metabolomicsultra-performance liquid chromatography-mass spectrometryblood metabolitesuntargeted metabolitestargeted metabolitesultra-performance liquid chromatography-mass spectrometryblood metabolitesuntargeted metabolitestargeted metabolites<p>Mass spectrometric analysis was performed using a Xevo TQ-XS mass spectrometer (Waters, USA) with an electrospray ionization (ESI) source operated in positive ion mode. Data were acquired in multiple reaction monitoring (MRM) mode using MassLynx software (v4.2). Instrument parameters included a source temperature of 150 °C, desolvation temperature of 500 °C, desolvation gas flow of 1000 L/h and capillary voltage of 3.0 kV.</p>TaurineHexanoylcarnitineDecanoylcarnitineOctanoylcarnitineAcetylcarnitineIsovaleryl L carnitineProlineHypoxanthinesn glycerol 3 glycerophosphocholineGlutamateHydroxypyridoxaminefalseIdentification and Validation of Plasma Lipid Biomarkers for Oral Cancer Diagnosis<p>Despite lipid metabolic changes in oral cancer (OC), their potential as diagnostic markers remains unclear. This study assessed the diagnostic potential of metabolites for OC detection. Plasma metabolites of patients with OC and healthy controls (HC) were profiled using untargeted and targeted metabolomics in discovery (182 OC, 364 HC) and external validation (52 OC, 52 HC) sets. Key biomarkers were selected via machine learning. Carnitine palmitoyltransferase 1 (CPT1) expression was assessed using enzyme-linked immunosorbent assay, immunohistochemistry, and public transcriptomic datasets. Functional assays based on CPT1 inhibition were conducted in OC cell lines. Machine learning identified OC diagnostic metabolites with high accuracy (AUC = 0.995) and targeted validation confirmed consistent trends. Three acylcarnitines—decanoyl-, octanoyl-, and hexanoylcarnitine—were downregulated in patient plasma, showing strong diagnostic performance (AUC = 0.941, 95% CI: 0.877–0.988) in external validation. Lipidomic and functional analyses revealed altered β-oxidation and glycerophospholipid metabolism, with CPT1 acting as a key regulator. CPT1 and acylcarnitines were abundant in OC tissues and cells. CPT1 inhibition suppressed OC cell growth and altered acylcarnitine levels. Our findings highlight three acylcarnitines as promising diagnostic biomarkers of OC and identify CPT1 as a central regulator of lipid metabolism, underscoring its relevance in OC pathogenesis and clinical application.</p>2026-05-022025-06-29MTBLS12655HMDB0000201HMDB0000688HMDB0000756HMDB0000705HMDB0000704HMDB0002110HMDB0000148HMDB0000254HMDB0000163HMDB0000194HMDB0000875