Analytic codes

## OverviewThis Zenodo record provides the R code used to reproduce the analyses for the manuscript:“Decision Support for Conditioning Intensity in Allogeneic Hematopoietic Cell Transplantation Using Predicted Survival”. The repository contains cohort construction, multiple imputation (MI), model development/validation, calibration/discrimination metrics, policy evaluation (TMLE), and sensitivity analyses (CIF analyses). Patient-level registry data are not included in this deposit. ## What is includedR scripts (organized as 001–014, including sub-numbered scripts) corresponding to the analytic workflow and figure/table generation. A README file describes the mapping between scripts and outputs. ## Data availabilityThis deposit does not include JSTCT/TRUMP or CIBMTR patient-level data. To run the scripts, users must supply the corresponding MI datasets (Excel files) with the expected sheet names/variables. Users should update file paths for their local environment as needed. ## How to runRun scripts in numerical order (001 → 014) unless otherwise noted in the README. Each script writes outputs to the directories specified in the script; output filenames include `run_date`. ## Software requirementsR (tested on R ≥ 4.2) and the required CRAN packages used in the scripts (e.g., survival, dplyr, tidyr, readxl, writexl, lspline, timeROC, etc.). Some scripts use parallel processing; set the number of cores appropriately for your environment.