MCLE-Mol: Empowering LLM with Molecular Comprehension and Low-Cost Continual Evolution for Interpretable Property Prediction

Large language models (LLMs) offer a new paradigm for molecular property prediction (MPP), yet a semantic gap between natural language and molecular representations limits LLMs' ability to capture structure-activity relationships (SAR). Recent approaches have explored injecting structure-level information into LLMs, primarily by modeling associations based on statistical regularities. However, these methods are prone to misinterpreting coincidental associations as general principles, imposing a bottleneck on predictive performance. To tackle the challenges above, we propose MCLE-Mol, an ML-LLM-Rule collaborative framework for MPP. It bridges the semantic gap by injecting ML-derived substructure attribution values into LLMs, utilizing Context-Calibrated Substructure Attribution Rules (CCSAR) to calibrate these attributions under specific chemical contexts to mitigate such misinterpretation. In addition, MCLE-Mol introduces a low-cost continual evolution strategy that updates CCSAR with frozen model parameters to adapt to dynamic chemical spaces. Experiments on multiple benchmark datasets demonstrate that MCLE-Mol outperforms all baselines, successfully resolving the trade-off between predictive performance and interpretability. Our code is available at https: //github.com/DoorKickers/MCLE-Mol . * Correctly Predicted Molecule * SMILES: smiles * Ground Truth Label: ground_truth_label * Model Prediction: predicted_label * SMILES_description: SMILES_description * SMILES_graph_description: SMILES_graph_description * SMILES_space_description: SMILES_space_description * Molecular_property: SMILES_property * valid_sub_smiles: valid_sub_smiles * valid_contributions: valid_contributions * Failed CoT Analysis: cot_analysis ## ORIGINAL MISPREDICTION DATA * Mispredicted Molecule * SMILES: smiles * Ground Truth Label: ground_truth_label * Model Prediction: predicted_label * SMILES_description: SMILES_description * SMILES_graph_description: SMILES_graph_description * SMILES_space_description: SMILES_space_description * Molecular_property: SMILES_property * valid_sub_smiles: valid_sub_smiles * valid_contributions: valid_contributions CORE PHILOSOPHY (Structure-State Logic): * Step 1: Check Presence: Does the molecule contain the specific substructure class relevant to the assay's biological mechanism? * Step 2: Describe State: If the substructure exists, you MUST generate a query describing its actual state (e.g., is the lipophilic core planar or twisted? is the electrophile steric-shielded or exposed?). Do not skip just because it looks "safe". * Step 3: Skip if Absent: Only skip a dimension if the molecule completely lacks the relevant substructure class. Note: The specific scenarios, biological targets, or mechanisms mentioned in any examples below are illustrative demonstrations only and are NOT exhaustive. Do not limit your analysis to these specific instances. You must derive the required structural state based on the actual chemical and physical reality of the specific molecule provided.