MEN2/RET Carrier Literature-Derived Benchmark Dataset for Medullary Thyroid Carcinoma Risk Stratification

This dataset release contains an anonymized, literature-derived benchmark dataset for exploratory machine-learning analysis of medullary thyroid carcinoma (MTC) status in Multiple Endocrine Neoplasia type 2 (MEN2) / RET-carrier records. The primary file, men2_ret_carrier_literature_dataset_v1.csv, contains 149 confirmed RET carriers curated from 10 peer-reviewed studies, spanning 14 RET variants. Extracted and harmonized variables include age, sex, RET variant, ATA-aligned risk level, calcitonin elevation and numeric calcitonin value, CEA numeric value, CEA imputation flag, thyroid nodule indicators, family history of MTC, pheochromocytoma, hyperparathyroidism, C-cell disease status, MEN2 syndrome status, and MTC diagnosis. The secondary file, men2_ret_carrier_synthetic_simulation_v1.csv, contains the case-control simulation artifact used for augmentation experiments in the associated reproducible machine-learning pipeline. It includes an is_synthetic flag and must not be treated as a real clinical cohort. This dataset is intended for research, reproducibility, education, and rare-disease benchmarking only. It is not a diagnostic, screening, triage, or clinical decision-support dataset and must not be used for real-world clinical decision-making. Because MEN2 is rare, users should not attempt to re-identify individuals or families from source-study, age, sex, genotype, or clinical-feature combinations. The associated reproducible code is available at https://github.com/arjuncodess/men2-predictor. Users should cite this dataset record and the original source studies listed in source_studies.csv.