Clinical and Serological Data for Sjögren's Syndrome Patients, 616 Cases

616 patient records from a multicenter study aim to identify the clinical signature of anti-centromere antibody-positive Sjögren’s syndrome. Wenlong Zhu published the data in 2026, using machine learning models like GBDT which achieved an AUC of 0.811. The analysis highlights predictors including serological markers, age, and Raynaud’s phenomenon.

Use Cases

• Train binary classification models to predict ACA-positive status based on clinical and serological features.
• Perform feature importance analysis using SHAP to understand key predictors like anti-SSA/Ro52 and age.
• Investigate non-linear variable interactions, such as how Raynaud’s phenomenon predictive value changes with age.
• Benchmark machine learning models (e.g., GBDT, LASSO) on a clinical cohort with a 7:3 train-validation split.

Strengths

• Includes data from 616 diagnosed patients across multiple clinical centers.
• Model validation was performed, with the GBDT model achieving an AUC of 0.811 in the validation cohort.
• Key predictive features are identified, including specific serological markers (anti-SSA/Ro52, anti-SSA/Ro60, anti-AMA-M2, anti-SSB, IgM), demographic factors, and Raynaud’s phenomenon.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small at 19.0 KB, indicating limited scope or a summary of findings rather than raw patient-level data.

Provenance

Source

Wenlong Zhu via figshare.

Collection Method

Multicenter study analyzing clinical data from a cohort of patients diagnosed with Sjögren’s syndrome.

Freshness

Last updated 2026-04-23 05:35:01; freshness should be verified.