DGSS: A Dynamic Interaction Graph Neural Network with Specific Substructure Awareness for Drug Synergy Prediction
Bibliographic record
Abstract
Combination therapy presents a transformative approach to treating complex diseases such as cancer by mitigating toxicity and resistance challenges inherent to monotherapy. A critical gap in current computational methods, however, lies in their inability to model cell-specific drug responses and dynamic drug-cell interactions, which are key factors in accurately predicting synergistic drug pairs. To address this, we propose DGSS, a novel Dynamic Interaction Graph Neural Network with Cell-Specific Drug Substructure Awareness, designed to explicitly capture two pivotal aspects: (1) drug substructures that drive efficacy in specific cellular environments, and (2) dynamic, context-dependent interactions between drugs and cell lines. Our framework introduces two technical innovations: a hierarchical attention mechanism that identifies cell-line-specific drug substructures by correlating molecular subgraphs with genomic features, and a dynamic graph network that models evolving cell-line states during drug exposure. Extensive experiments under three data partitioning strategies across 12 datasets demonstrate DGSS's robustness, consistently outperforming all state-of-the-art baseline models. On the Loewe Synergy dataset, the model achieved AUROC and AUPRC of 96.0% and 85.5%, respectively, and exhibited good stability. By bridging molecular substructure dynamics with cellular context, DGSS advances precision in synergy prediction, offering a data-driven framework to optimize combination therapies in personalized oncology.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".