{"slug": "smilesgnn-interpretable-clinical-toxicity-prediction-via-smiles-graph-cross", "title": "SMILESGNN: Interpretable Clinical Toxicity Prediction via SMILES-Graph Cross-Attention Fusion", "summary": "SMILESGNN, a multimodal architecture that fuses a SMILES Transformer encoder with a GATv2 graph encoder via cross-attention, achieved AUC-ROC 0.987 and F1 0.906 on ClinTox with only 0.4M parameters, according to an arXiv paper (2609.28553v1). The SMILESGNN-PT variant, which uses a ChemBERTa-2 pretrained backbone, reached a mean AUC-ROC of 0.750 across the 12 tasks of Tox21, comparable to ChemBERTa-2 alone and a same-backbone concat-fusion baseline. The authors report that cross-attention is a practical fusion alternative that preserves competitive predictive performance while retaining an explicit graph branch for GNNExplainer-based substructure analysis of toxic predictions.", "body_md": "arXiv:2609.28553v1 Announce Type: new \nAbstract: Drug toxicity prediction is critical for reducing late-stage attrition in drug discovery, yet remains challenging due to severe class imbalance, scaffold-based generalization, and the clinical need for interpretable predictions. Single-modality approaches-SMILES Transformers or graph neural networks capture complementary aspects of molecular structure, while sequence-only models cannot directly provide graph-attributed explanations. We present SMILESGNN, a multimodal architecture that fuses a SMILES Transformer encoder and a GATv2 graph encoder via cross-attention, and SMILESGNN-PT, a variant using a ChemBERTa-2 pretrained backbone. The design retains an explicit graph branch within the predictive pipeline, supporting GNNExplainer-based analysis of substructures associated with toxic predictions. On ClinTox, SMILESGNN achieves AUC-ROC 0.987 and F1 0.906 with only 0.4M parameters, performing competitively with a strong SMILESTransformer and a larger ChemBERTa-2/GATv2 concat-fusion baseline. On Tox21 (12 tasks), SMILESGNN-PT obtains mean AUC-ROC 0.750, comparable to ChemBERTa-2 alone and the same-backbone concat-fusion baseline. Overall, the results suggest that cross-attention is a practical fusion alternative that preserves competitive predictive performance while enabling graph-based interpretability support.", "url": "https://wpnews.pro/news/smilesgnn-interpretable-clinical-toxicity-prediction-via-smiles-graph-cross", "canonical_source": "https://arxiv.org/abs/2609.28553", "published_at": "2026-09-25 04:00:00+00:00", "updated_at": "2026-09-25 04:30:00.152646+00:00", "lang": "en", "topics": ["machine-learning", "ai-research", "neural-networks", "ai-safety"], "entities": ["SMILESGNN", "SMILESGNN-PT", "ChemBERTa-2", "GATv2", "GNNExplainer", "ClinTox", "Tox21", "arXiv"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/smilesgnn-interpretable-clinical-toxicity-prediction-via-smiles-graph-cross", "markdown": "https://wpnews.pro/news/smilesgnn-interpretable-clinical-toxicity-prediction-via-smiles-graph-cross.md", "text": "https://wpnews.pro/news/smilesgnn-interpretable-clinical-toxicity-prediction-via-smiles-graph-cross.txt", "jsonld": "https://wpnews.pro/news/smilesgnn-interpretable-clinical-toxicity-prediction-via-smiles-graph-cross.jsonld"}}