Introduction
Predicting and understanding the heterogeneity of drug responses is crucial for precision oncology. In this study, we introduce a multi-contextual self-alignment (MCSA) framework that integrates semantic knowledge of pharmacogenomic data, facilitates cross-domain interactions across heterogeneous drugs, transcriptomic profiles, and multi-institutional datasets, and enables genomic interpretation of both targeted therapies and chemotherapeutic agents. The model is designed to function effectively in both dynamic and static data environments, exhibiting adaptability to diverse drugs, cancer types, and sequencing technologies. By leveraging genomic information, MCSA elucidates drug mechanisms of action through cellular state analysis and assesses the efficacy of clinical chemotherapy. Its universality and relevance to clinical outcomes have been validated through applications across various methods and clinical datasets. Overall, MCSA offers a unified framework for precision oncology and drug discovery, underscoring its potential to advance the life sciences.
Figure 1. The model architecture of MCSA. MCSA is a continual learning framework designed to adopt changing dynamics in drug response data, and it can be mainly divided into three parts: local alignment interpretability alignment, and global alignment.