Peeling Context from Cause for Multimodal Molecular Property Prediction
This addresses the need for more reliable and interpretable predictors in molecular design, though it is incremental as it builds on existing multimodal and causal methods.
The paper tackled the problem of deep models relying on spurious context rather than causal structure in molecular property prediction, which reduces reliability under distribution shift. The result was that CLaP consistently improved MAE, MSE, and R² over competitive baselines across four molecular benchmarks.
Deep models are used for molecular property prediction, yet they are often difficult to interpret and may rely on spurious context rather than causal structure, which reduces reliability under distribution shift and harms predictive performance. We introduce CLaP (Causal Layerwise Peeling), a framework that separates causal signal from context in a layerwise manner and integrates diverse graph representations of molecules. At each layer, a causal block performs a soft split into causal and non-causal branches, fuses causal evidence across modalities, and progressively removes batch-coupled context to focus on label-relevant structure, thereby limiting shortcut signals and stabilizing layerwise refinement. Across four molecular benchmarks, CLaP consistently improves MAE, MSE, and $R^2$ over competitive baselines. The model also produces atom-level causal saliency maps that highlight substructures responsible for predictions, providing actionable guidance for targeted molecular edits. Case studies confirm the accuracy of these maps and their alignment with chemical intuition. By peeling context from cause at every layer, the model yields predictors that are both accurate and interpretable for molecular design.