For patients in the intervention group, the daGOAT model will be used to predict the occurrence of severe acute GvHD from day 17 to day 23 after transplantation. Each subject will be stratified into low-, intermediate-, and high-risk groups, and corresponding preventive interventions were implemented according to their risk levels.

For patients in the control group, they will receive standard prophylaxis without additional GvHD prophylactic agents, including mesenchymal cell infusion, anti-CD25 monoclonal antibodies, and ruxolitinib outside the scope specified in the study protocol.

1. Model-predicted high-risk patients: will receive standard prophylaxis plus ruxolitinib 5mg twice daily (bid) orally until at least day 60 post-transplantion and will be terminated after day 100. If severe hematological signs occur such as when there is severe neutropenia (<0.1×10^9/L), ruxolitinib can be used at half dose or discontinued until recovery.
2. Model-predicted moderate-risk patients: will receive standard prophylaxis plus ruxolitinib 5mg once daily (qd) orally until at least day 60 post-transplantion and will be terminated after day 100. If severe hematological signs occur such as when there is severe neutropenia (<0.1×10^9/L), ruxolitinib can be used at half dose or discontinued until recovery.
3. Model-predicted low-risk patients: will receive standard prophylaxis without additional GvHD prophylactic agents, including mesenchymal cell infusion, anti-CD25 monoclonal antibodies, and ruxolitinib outside the scope specified in the study protocol.

Incidence of severe aGVHD after transplantation within 100 days. The medical records for each case wil be reviewed by two or three physicians to confirm the aGVHD diagnosis and grading (according to the MAGIC criteria)