{
  "abstract": "Blood-based metabolomic profiling has been widely investigated for breast cancer (BC) detection; however, clinical implementation remains limited due to variability in sample handling, analytical reproducibility, and overfitting during statistical analysis. We established a plasma GC/MS metabolomics workflow for discriminating BC from healthy controls (HC) using conventional machine-learning algorithms. Plasma samples (n = 360; BC = 180, HC = 180) were collected prospectively under standardized preanalytical conditions before surgery and the initiation of systematic anticancer therapy and analyzed using a quantitative GC/MS platform with automated derivatization. Feature selection and model development were conducted using three machine-learning (ML) algorithms (Lasso logistic regression (LR), random forest classifier (RFC), and support vector machine (SVM)). A total of 45 metabolite candidate biomarkers were identified, and the optimal number of metabolite features for each algorithm was estimated by a recursive feature elimination (RFE)-based strategy. The best-performing models achieved area under the ROC curve values (AUC) of 0.910 (LR), 0.893 (RFC), and 0.843 (SVM). We selected prioritizing candidate biomarkers consistently expressed across the multi-algorithm pipeline. A bagging ensemble model improved stability (AUC = 0.911) and reduced false-positive predictions in the independent HC dataset. In addition, model stability with respect to false-positive predictions was assessed using an independent HC cohort (n = 15) that was collected at a separate institution. These results indicate that a plasma metabolomics workflow combined with conventional multi-algorithm ML, algorithm-specific feature selection, and independent assessment provides stable discrimination between BC and HC in a moderately sized cohort.",
  "abstract_shown": true,
  "affiliations": [
    "MS Business Unit"
  ],
  "article_type": "New results",
  "arxiv": null,
  "author_total": 14,
  "authors": [
    {
      "affs": [
        0
      ],
      "name": "N. Ojima"
    },
    {
      "name": "T. Yokoyama"
    },
    {
      "name": "S. Nishiumi"
    },
    {
      "name": "S. Kikuchi"
    },
    {
      "name": "K. Yoneda"
    },
    {
      "name": "M. Kimura"
    },
    {
      "name": "K. Tanaka"
    },
    {
      "name": "M. Morimoto"
    },
    {
      "name": "A. Miyazaki"
    },
    {
      "name": "T. Tomita"
    },
    {
      "name": "M. Ohyanagi"
    },
    {
      "name": "M. Nagahashi"
    },
    {
      "name": "Y. Miyoshi"
    },
    {
      "name": "M. Ikeda"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/2dg2r54v7xmvr.bib",
  "cited_by": 0,
  "doi": "10.64898/2026.09.23.753710",
  "fulltext_url": "https://www.biorxiv.org/content/10.64898/2026.09.23.753710v1.full",
  "id": "2dg2r54v7xmvr",
  "issns": [],
  "landing_url": "https://www.biorxiv.org/content/10.64898/2026.09.23.753710v1",
  "language": null,
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "biorxiv",
      "cancer biology"
    ],
    [
      "openalex",
      "1312|Metabolomics and Mass Spectrometry Studies"
    ]
  ],
  "openalex": "W7214719894",
  "pdf_url": "https://www.biorxiv.org/content/10.64898/2026.09.23.753710v1.full.pdf",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-23",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/2dg2r54v7xmvr.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:25:00Z",
      "native_id": "10.64898/2026.09.23.753710",
      "source": "biorxiv",
      "url": "https://www.biorxiv.org/content/10.64898/2026.09.23.753710v1"
    }
  ],
  "subjects": [
    {
      "field": "medicine",
      "origin": "biorxiv:cancer biology",
      "subfield": "oncology"
    }
  ],
  "title": "A plasma metabolomics workflow for breast cancer detection using quantitative GC/MS and machine learning",
  "title_text": "A plasma metabolomics workflow for breast cancer detection using quantitative GC/MS and machine learning",
  "type": "preprint",
  "updated_at": "2026-10-30T12:25:00Z",
  "url": "https://research.draconitic.com/paper/2dg2r54v7xmvr",
  "venue": "bioRxiv",
  "versions": []
}