![[2601.11283] Metabolomic Biomarker Discovery for ADHD Diagnosis Using Interpretable Machine Learning](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2601.11283] Metabolomic Biomarker Discovery for ADHD Diagnosis Using Interpretable Machine Learning
Summary
This article presents a novel approach to ADHD diagnosis by integrating urinary metabolomics with interpretable machine learning, identifying key biochemical signatures for more objective assessments.
Why It Matters
ADHD diagnosis currently lacks objective tools, making this research significant as it combines metabolomics with machine learning to provide a potential framework for precision psychiatry. The findings could lead to improved diagnostic methods and better understanding of ADHD's biological underpinnings.
Key Takeaways
- The study identifies 14 metabolites linked to ADHD, enhancing diagnostic accuracy.
- A Closest Resemblance classifier outperformed traditional methods, achieving an AUC > 0.97.
- The research highlights the importance of integrating biological data with machine learning for mental health diagnostics.
- Transparent decision-making in the model supports its application in clinical settings.
- Findings may pave the way for future point-of-care diagnostic tools.
Computer Science > Machine Learning arXiv:2601.11283 (cs) [Submitted on 16 Jan 2026 (v1), last revised 24 Feb 2026 (this version, v2)] Title:Metabolomic Biomarker Discovery for ADHD Diagnosis Using Interpretable Machine Learning Authors:Nabil Belacel, Mohamed Rachid Boulassel View a PDF of the paper titled Metabolomic Biomarker Discovery for ADHD Diagnosis Using Interpretable Machine Learning, by Nabil Belacel and Mohamed Rachid Boulassel View PDF HTML (experimental) Abstract:Attention Deficit Hyperactivity Disorder (ADHD) is a prevalent neurodevelopmental disorder with limited objective diagnostic tools, highlighting the urgent need for objective, biology-based diagnostic frameworks in precision psychiatry. We integrate urinary metabolomics with an interpretable machine learning framework to identify biochemical signatures associated with ADHD. Targeted metabolomic profiles from 52 ADHD and 46 control participants were analyzed using a Closest Resemblance (CR) classifier with embedded feature selection. The CR model outperformed Random Forest and K-Nearest Neighbor classifiers, achieving an AUC > 0.97 based on a reduced panel of 14 metabolites. These metabolites including dopamine 4-sulfate, N-acetylaspartylglutamic acid, and citrulline map to dopaminergic neurotransmission and amino acid metabolism pathways, offering mechanistic insight into ADHD pathophysiology. The CR classifier's transparent decision boundaries and low computational cost support integration into targe...