The paper introduces a fragment‑level graph encoding for molecular property prediction, using BRICS decomposition to represent molecules as graphs of synthetically meaningful fragments. This change of input representation allows any message‑passing GNN and post‑hoc attribution method to produce explanations at the functional‑group level, matching atom‑level accuracy on smaller graphs. Experiments across multiple architectures, tasks, and explainers show that fragment‑level encoding recovers known pharmacophores more reliably than post‑hoc aggregation of atom‑level explanations.
arXiv:2608.30636v1 Announce Type: new Abstract: Optimizing absorption, distribution, metabolism, and excretion (ADME) is an important part of small molecule drug discovery. Many machine learning mode...
arXiv:2607. 02212v1 Announce Type: cross Abstract: Aqueous solubility is a key property in early-stage drug discovery, but most predictive models merge physicochemical descriptors and molecular graph information into a single representation, obscuring whether a prediction is driven by global chemistry, molecular structure, or both.
WOMBAT is a benchmark comprising 14 whitebox graph neural networks (GNNs) whose message‑passing weights are manually set to detect specific SMARTS motifs. Each model’s decision rule is explicitly known, providing a ground truth for attribution that allows researchers to identify and study errors in post‑hoc explainers such as GNNExplainer, PGExplainer, and Integrated Gradients. The authors validate the models on millions of PubChem molecules, demonstrate how Integrated Gradients can be misled to spread attribution, and release the dataset, models, and evaluation code for future XAI tool development.
SMILESGNN is a multimodal architecture that fuses a SMILES Transformer encoder with a GATv2 graph encoder through cross‑attention, enabling interpretable clinical toxicity predictions. The model retains an explicit graph branch, allowing GNNExplainer to identify substructures linked to toxicity. On the ClinTox dataset it achieves an AUC‑ROC of 0.987 and F1 of 0.906 with only 0.4 M parameters, while on Tox21 it attains a mean AUC‑ROC of 0.750, comparable to strong single‑modality baselines.
The paper introduces MPGE, a Multi‑Perspective Graph Explainer that provides three complementary explanations—factual support, counterfactual sensitivity, and exemplar tolerance—for graph neural network predictions on molecular data. It formulates a shared constrained optimization that balances prediction preservation, compactness, and edit cost, and extends CF‑GNNExplainer to learn symmetric edge rankings for counterfactuals. Experiments on several benchmark datasets show that factual masks retain a small fraction of edges while counterfactual and exemplar explanations achieve high coverage, demonstrating the method’s ability to reveal model behavior without claiming chemical validity.
arXiv:2604. 06336v2 Announce Type: replace-cross Abstract: Fragment-level representations provide a natural way to capture recurring molecular substructures and reuse their learned representations across molecules.
arXiv:2602. 20573v3 Announce Type: replace Abstract: Molecules are often represented as SMILES strings, which can be readily converted to hand-crafted descriptors or fingerprints (FP) for molecular property prediction.
arXiv:2609.37555v1 Announce Type: new Abstract: Drug discovery is a costly and high-risk process, where toxicity-related failures remain a major cause of attrition in both preclinical and clinical st...
arXiv:2607. 07935v1 Announce Type: cross Abstract: We present path_boost, a Python package for interpretable supervised learning on graph-structured input data.
arXiv:2608. 10480v1 Announce Type: new Abstract: Large language models (LLMs) are widely applied across chemical tasks, such as molecular property prediction, which underpins drug discovery.
arXiv:2511. 19264v2 Announce Type: replace-cross Abstract: Generative Flow Networks (GFlowNets) construct molecules through sequential decisions, but their internal policies remain opaque, limiting adoption in drug discovery, where chemists need interpretable rationales for proposed structures.
arXiv:2609.05694v1 Announce Type: new Abstract: Predicting olfactory qualities from molecular structure is an open problem in chemoinformatics. Although linear models can link molecular features to o...