Oral Presentation Accepted at CIBB 2026

I am thrilled to announce that our short paper “Dual-view Atomic and Fragment Graph Representation Learning for Drug Response Prediction” has been accepted for an ORAL PRESENTATION at CIBB 2026 (the 21st International Conference on Computational Intelligence Methods for Bioinformatics and Biostatistics), held at Sapienza University of Rome, Italy, from September 2–4, 2026.

Out of a record number of submissions this year, we are honored that our work was selected among the few contributions chosen for oral presentation. The paper proposes DAF-DRP, a dual-view GNN framework that jointly models molecules through both atomic-level and fragment-level graph representations, integrated with gene expression profiles via multiple fusion strategies. Our experiments on CCLE and gCSI datasets demonstrate consistent improvements over established baselines including GraphDRP, DeepCDR, HiDRA, and UNO.

We received encouraging feedback from the reviewers, particularly noting that “the proposed approach presents a promising and well-motivated contribution for incorporating higher-order chemical semantics into drug response prediction.” We are now preparing the camera-ready version addressing reviewer comments on methodological details, ablation studies, and statistical significance testing, and look forward to presenting in Rome this September.

This work represents an important step in my ongoing research at the intersection of AI, computational biology, and biophysics — exploring how multi-scale molecular representations can improve predictive modeling in precision medicine.