.png){kind=icon}
.png){kind=icon}
Return
TAPB: an interventional debiasing framework for alleviating target prior bias in drug-target interaction prediction
.png){kind=icon}
OA
AI
DOI：10.1038/s41467-025-66915-1.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Drug Target Interaction (DTI) prediction is vital for drug repurposing. Previous DTI studies on BioSNAP and BindingDB datasets often attribute biased predictions to “drug bias,” while our work reveals “target prior bias” as the predominant issue. This bias stems from the “prior tendency,” characterized by the imbalanced label distribution of targets in the training data. From causal lens, target “prior tendency” is a confounder, causing models trained with P(Y∣D, T) to learn spurious associations between targets and labels rather than genuine interaction mechanisms. In this study, we introduce alleviating Target Prior Bias in Drug-Target Interaction Prediction (TAPB), a novel debiasing framework that employs amino acid randomization, confounder alignment module (CAM), and interventional training to compute P(Y∣D, do(T)) via backdoor adjustment, thereby addressing this bias. TAPB achieves competitive performance over existing approaches, demonstrating enhanced generalization and providing interpretable insights into DTIs. In this work, authors show that while previous Drug-Target Interaction prediction studies attribute biased predictions to drug bias, they reveal target prior bias as the key issue arising from imbalanced training data. They develop TAPB which is a debiasing framework that mitigates target prior bias via causal debiasing, achieving stronger generalization and interpretability.
Keywords:
Drug-Target Interaction
Target Prior Bias
Causal Debiasing
Interventional Training
Confounder Alignment
Journal
IF:
15.7
Papers:
9.2W
Citations:
91.2W
View more organizations