Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 63 of the 64 references it cites were found in a registry or an index; 19 of the 30 sources its claims cite were checked for support.

Identification of novel diagnostic panel for mild cognitive impairment and Alzheimer’s disease: findings based on urine proteomics and machine learning

Open access
Wang, Yuye, Sun, Yu, Wang, Yu, Jia, Shuhong, Qiao, Yanan, Zhou, Zhi, Shao, Wen, Zhang, Xiangfei, Guo, Jing, Zhang, Bin, Niu, Xiaoqian, Wang, Yi, Peng, Dantao · 2023 · Alzheimer's Research & Therapy

Abstract

Background: Alzheimer’s disease is a prevalent disease with a heavy global burden. Proteomics is the systematic study of proteins and peptides to provide comprehensive descriptions. Aiming to obtain a more accurate and convenient clinical diagnosis, researchers are working for better biomarkers. Urine is more convenient which could reflect the change of disease at an earlier stage. Thus, we conducted a cross-sectional study to investigate novel diagnostic panels. Methods: We firstly enrolled participants from China-Japan Friendship Hospital from April 2022 to November 2022, collected urine samples, and conducted an LC–MS/MS analysis. In parallel, clinical data were collected, and clinical examinations were performed. After statistical and bioinformatics analyses, significant risk factors and differential urinary proteins were determined. We attempt to investigate diagnostic panels based on machine learning including LASSO and SVM. Results: Fifty-seven AD patients, 43 MCI patients, and 62 CN subjects were enrolled. A total of 3366 proteins were identified, and 608 urine proteins were finally included in the analysis. There were 33 significantly differential proteins between the AD and CN groups and 15 significantly differential proteins between the MCI and CN groups. AD diagnostic panel included DDC, CTSC, EHD4, GSTA3, SLC44A4, GNS, GSTA1, ANXA4, PLD3, CTSH, HP, RPS3, CPVL, age, and APOE ε4 with an AUC of 0.9989 in the training test and 0.8824 in the test set while MCI diagnostic panel included TUBB, SUCLG2, PROCR, TCP1, ACE, FLOT2, EHD4, PROZ, C9, SERPINA3, age, and APOE ε4 with an AUC of 0.9985 in the training test and 0.8143 in the test set. Besides, diagnostic proteins were weakly correlated with cognitive functions. Conclusions: In conclusion, the procedure is convenient, non-invasive, and useful for diagnosis, which could assist physicians in differentiating AD and MCI from CN.

Overview

CoChat Summary

TL;DR

This cross-sectional urine proteomics and machine-learning study identified candidate diagnostic panels that distinguished AD and MCI from cognitively normal controls, with test-set AUCs of 0.8824 for AD and 0.8143 for MCI. The panels combined urinary proteins with age and APOE ε4 status, though diagnostic proteins correlated only weakly with cognitive tests.

Key findings

  • In 57 AD, 43 MCI, and 62 CN participants, LC-MS/MS identified 3,366 urine proteins; 608 detected in >50% of samples were analyzed. There were 33 differential proteins between AD and CN (21 upregulated, 12 downregulated) and 15 between MCI and CN (7 upregulated, 8 downregulated).
  • The AD diagnostic panel included DDC, CTSC, EHD4, GSTA3, SLC44A4, GNS, GSTA1, ANXA4, PLD3, CTSH, HP, RPS3, CPVL, age, and APOE ε4; it achieved an AUC of 0.8824 in the test set.
  • The MCI diagnostic panel included TUBB, SUCLG2, PROCR, TCP1, ACE, FLOT2, EHD4, PROZ, C9, SERPINA3, age, and APOE ε4; it achieved an AUC of 0.8143 in the test set.
  • EHD4 was selected for both panels; GSTA1 was downregulated in both AD and MCI, while EHD4 and C9 were upregulated in both. Diagnostic proteins were mostly weakly correlated with cognitive tests; HP, TUBB, and FLOT2 showed no significant correlations.

Methods

  • Cross-sectional study at China-Japan Friendship Hospital (April–November 2022) enrolling 162 participants >50 years: 57 AD, 43 MCI, 62 CN. Clinical history, neuropsychological battery, APOE genotype, and qEEG/MRI in most were collected; midstream urine was analyzed by LC-MS/MS in DDA mode.
  • Label-free quantification used iBAQ/FOT; proteins detected in >50% of samples were retained (608), with missing values imputed by KNN. Differential proteins used limma with p < 0.05 and |log2FC| > 0.58; PPI, GO, and GSEA analyses were performed.
  • LASSO selected proteins plus age and APOE ε4; SVM classifiers with tenfold cross-validation compared linear, polynomial, radial, and sigmoid kernels; data were split 0.7 training/0.3 test, with accuracy and AUC evaluated.

Limitations

  • Paper-stated: single-site sample with no multi-hospital/community real-world validation; relatively few proteins detected in >50% of samples; no in vivo or in vitro experiments on mechanisms; machine-learning feature selection used differential proteins derived from the whole dataset, so test-set performance may be optimistic and some results may be coincidental.
  • Evident from the text: the cross-sectional design cannot establish temporal or causal relationships; the sample size is modest and groups differ in age and APOE ε4 distribution, which may affect comparisons.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1186/s13195-023-01324-4

Cites

63 of 63 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1 Differential urinary proteins in the AD and MCI groups compared to CN. A Heatmap of a total of 33 differential proteins between AD and CN. B Volcano plot showed the distribution of all proteins between AD and CN. The red dots were coincident with the left heatmap. C Heatmap of a total of 15 differential proteins between MCI and CN. D Volcano plot shows the distribution of all proteins between AD and CN. The red dots were coincident with the left heatmap. The horizontal dashed line indicates the threshold of p-value (− log100.05 ≈ 1.3). The vertical dashed line indicates the threshold of fold change (± log21.5 ≈ ± 0.58)
Fig. 1 Differential urinary proteins in the AD and MCI groups compared to CN. A Heatmap of a total of 33 differential proteins between AD and CN. B Volcano plot showed the distribution of all proteins between AD and CN. The red dots were coincident with the left heatmap. C Heatmap of a total of 15 differential proteins between MCI and CN. D Volcano plot shows the distribution of all proteins between AD and CN. The red dots were coincident with the left heatmap. The horizontal dashed line indicates the threshold of p-value (− log100.05 ≈ 1.3). The vertical dashed line indicates the threshold of fold change (± log21.5 ≈ ± 0.58)CC BY · PubMed Central
Fig. 2 Protein–protein interaction network of significantly differential proteins. A Network of AD-CN differential proteins. B Network of MCI-CN differential proteins. The size of the node indicated relative fold change of differential proteins when compared to the controls. Red indicated upregulation, and blue indicated downregulation
Fig. 2 Protein–protein interaction network of significantly differential proteins. A Network of AD-CN differential proteins. B Network of MCI-CN differential proteins. The size of the node indicated relative fold change of differential proteins when compared to the controls. Red indicated upregulation, and blue indicated downregulationCC BY · PubMed Central
Fig. 3 Diagnostic panel constructed by LASSO model. A LASSO model for variable selection in the AD-CN group. B Boxplot of the included diagnostic proteins for AD diagnosis. C LASSO model for variable selection in the MCI-CN group. D Boxplot of the included diagnostic proteins for MCI diagnosis. *p < 0.05; **p < 0.01. DDC, dopa decarboxylase; CTSC, cathepsin C; EHD4, EH domain containing 4; GSTA3, glutathione S-transferase alpha 3; SLC44A4, solute carrier family 44 member 4; GNS, glucosamine (N-acetyl)-6-sulfatase; GSTA1, glutathione S-transferase alpha 1; ANXA4, annexin A4; PLD3, phospholipase D family member 3; CTSH, cathepsin H; HP, haptoglobin; RPS3, ribosomal protein S3; CPVL, carboxypeptidase vitellogenic like; TUBB, tubulin beta class I; SUCLG2, succinate-CoA ligase GDP-forming subunit beta; PROCR, protein C receptor; TCP1, T-complex 1; ACE, angiotensin I-converting enzyme; FLOT2, flotillin 2; PROZ, protein Z, vitamin K-dependent plasma glycoprotein; C9, complement C9; SERPINA3, serpin family A member 3
Fig. 3 Diagnostic panel constructed by LASSO model. A LASSO model for variable selection in the AD-CN group. B Boxplot of the included diagnostic proteins for AD diagnosis. C LASSO model for variable selection in the MCI-CN group. D Boxplot of the included diagnostic proteins for MCI diagnosis. *p < 0.05; **p < 0.01. DDC, dopa decarboxylase; CTSC, cathepsin C; EHD4, EH domain containing 4; GSTA3, glutathione S-transferase alpha 3; SLC44A4, solute carrier family 44 member 4; GNS, glucosamine (N-acetyl)-6-sulfatase; GSTA1, glutathione S-transferase alpha 1; ANXA4, annexin A4; PLD3, phospholipase D family member 3; CTSH, cathepsin H; HP, haptoglobin; RPS3, ribosomal protein S3; CPVL, carboxypeptidase vitellogenic like; TUBB, tubulin beta class I; SUCLG2, succinate-CoA ligase GDP-forming subunit beta; PROCR, protein C receptor; TCP1, T-complex 1; ACE, angiotensin I-converting enzyme; FLOT2, flotillin 2; PROZ, protein Z, vitamin K-dependent plasma glycoprotein; C9, complement C9; SERPINA3, serpin family A member 3CC BY · PubMed Central
Fig. 4 ROC curve for AD and MCI diagnosis in different SVM models. A ROC curve for AD diagnosis in the training set. B ROC curve for AD diagnosis in the test set. C ROC curve for MCI diagnosis in the training set. D ROC curve for MCI diagnosis in the test set
Fig. 4 ROC curve for AD and MCI diagnosis in different SVM models. A ROC curve for AD diagnosis in the training set. B ROC curve for AD diagnosis in the test set. C ROC curve for MCI diagnosis in the training set. D ROC curve for MCI diagnosis in the test setCC BY · PubMed Central
Fig. 5 Correlation heatmap between diagnostic proteins and cognition tests. *p < 0.05; **p < 0.01; ***p < 0.001
Fig. 5 Correlation heatmap between diagnostic proteins and cognition tests. *p < 0.05; **p < 0.01; ***p < 0.001CC BY · PubMed Central

Verification

Claims the paper makes

18 Valid · 0 Not valid · 8 Not checked · 45 Own results and interpretations

20 more claims

Create a free account to see all 71 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 40
  • It may state two separate claims as one. 15

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

Show all 64 references

Cited by

See all 34

Metadata

Authors
Wang, Y., Sun, Y., Wang, Y., Jia, S., Qiao, Y., Zhou, Z., Shao, W., Zhang, X., Guo, J., Zhang, B., Niu, X., Wang, Y., & Peng, D.
Affiliations
  • Chinese Academy of Medical Sciences & Peking Union Medical College (CN)
  • China-Japan Friendship Hospital (CN)
  • Peking University (CN)
Corresponding author
  • Wang, Yu
  • Wang, Yi
  • Peng, Dantao
Venue
Alzheimer's Research & Therapy
Published
2023-11-04
Volume
15
Issue
1
Pages
191-191
ISSN
1758-9193 (ISSN-L)
OpenAlex
W4388341887
Cited by
34
References
64
Citations by year
2026: 7 · 2025: 19 · 2024: 8
FWCI
6.89
Type
journal article
Language
en
Open access
Licence
Locations
  • Alzheimer s Research & Therapy (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
National Key Research and Development Program of China (2018YFA0507503, 2022YFC2010103)
Topics
  • Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
  • Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Alzheimer's disease research and treatments
  • Physical Sciences › Chemistry › Spectroscopy › Advanced Proteomics Techniques and Applications
Keywords
Alzheimer's disease, mild cognitive impairment, urine proteomics, diagnostic biomarkers, diagnostic panel, APOE ε4, non-invasive diagnosis, machine learning classification, LASSO, support vector machine
MeSH terms
Machine Learning, Alzheimer Disease, Chromatography, Liquid, Cross-Sectional Studies, Humans, Biomarkers, Proteomics, Apolipoprotein E4, Tandem Mass Spectrometry, Cognitive Dysfunction

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-16, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-16 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Identification of novel diagnostic panel for mild cognitive impairment and Alzheimer’s disease: findings based on urine proteomics and machine learning. https://app.cochat.ai/papers/10.1186/s13195-023-01324-4
The paper
Wang, Y., Sun, Y., Wang, Y., Jia, S., Qiao, Y., Zhou, Z., Shao, W., Zhang, X., Guo, J., Zhang, B., Niu, X., Wang, Y., & Peng, D. (2023). Identification of novel diagnostic panel for mild cognitive impairment and Alzheimer’s disease: findings based on urine proteomics and machine learning. Alzheimer's Research & Therapy. https://doi.org/10.1186/s13195-023-01324-4

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.