Data Sheet 1_Identifying the clinical signature of anti-centromere antibody-positive Sjögr

616 patients diagnosed with Sjögren's syndrome, comprising 81 anti-centromere antibody (ACA)-positive and 535 ACA-negative cases, were analyzed in a multicenter study. The dataset was used to develop a machine learning model, with a Gradient Boosted Decision Tree (GBDT) achieving an AUC of 0.811 in validation. The research was authored by Wenlong Zhu and last updated on the figshare platform in April 2026.

Use Cases

• Predicting anti-centromere antibody (ACA) positivity in Sjögren's syndrome patients based on clinical and serological markers.
• Analyzing feature importance for ACA status using SHAP values, focusing on predictors like anti-SSA/Ro52, age, and Raynaud's phenomenon.
• Investigating non-linear interactions between clinical variables, such as how the predictive contribution of Raynaud's phenomenon changes with age.
• Comparing the performance of different machine learning models (e.g., LASSO, GBDT) on a clinical classification task.

Strengths

• Data originates from a multicenter cohort of 616 patients, which may improve generalizability compared to single-center studies.
• Model development used a 7:3 train-validation split and reported specific performance metrics, including an AUC of 0.811 with a 95% confidence interval.
• The study employed SHAP analysis to quantify and interpret the influence of individual features like serological markers and demographic factors.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the underlying source data is unknown, which may limit suitability assessment for specific modeling tasks.
• The dataset is very small (9.0 KB), suggesting it may contain only summary or model results rather than the full patient-level data.

Provenance

Source

figshare, author Wenlong Zhu.

Collection Method

Clinical data was collected from a multicenter cohort study of patients diagnosed with Sjögren's syndrome.

Freshness

Last updated 2026-04-23 05:34:59; freshness should be verified.