InstructMol: Multi-Modal Molecular Assistant

Paper Information

Citation: Cao, H., Liu, Z., Lu, X., Yao, Y., & Li, Y. (2025). InstructMol: Multi-Modal Integration for Building a Versatile and Reliable Molecular Assistant in Drug Discovery. Proceedings of the 31st International Conference on Computational Linguistics, 354-379.

Publication: COLING 2025

Additional Resources:

• Official Repository

What kind of paper is this?

Methodological Paper ($\Psi_{\text{Method}}$)

This work proposes InstructMol, a novel multi-modal architecture and training paradigm. It focuses on engineering a system that aligns a pre-trained molecular graph encoder with a general-purpose Large Language Model (LLM). The paper’s primary contribution is the Two-Stage Instruction Tuning strategy (Alignment Pre-training + Task-Specific Tuning) designed to bridge the modality gap between 2D molecular graphs and natural language.

What is the motivation?

Current AI approaches in drug discovery face a dichotomy:

1. Specialist Models: Highly effective at specific tasks (e.g., property prediction) but lack adaptability and require extensive labeled data.
2. Generalist LLMs: Capable of reasoning and dialogue but struggle to process complex scientific data like molecular structures natively, often relying on text representations (SMILES) which can be semantically brittle.

There is a practical need for a unified “Molecular Assistant” that can visually understand molecular structures (via graphs), reason about them in natural language, and generalize across diverse tasks like synthesis planning and property analysis without training from scratch.

What is the novelty here?

The core novelty lies in the architecture and the two-stage training pipeline designed to align differing modalities efficiently:

1. MoleculeSTM Integration: Unlike prior works using GraphMVP, InstructMol initializes its graph encoder with MoleculeSTM, which is already pre-aligned with text via contrastive learning, facilitating easier downstream alignment.
2. Two-Stage Alignment Strategy:
   • Stage 1 (Alignment Pre-training): Freezes both the LLM and Graph Encoder; trains only a linear projector using a massive dataset of molecule-description pairs to map graph features into the LLM’s token space.
   • Stage 2 (Task-Specific Instruction Tuning): Freezes the Graph Encoder; fine-tunes the Projector and the LLM (using LoRA) on specific downstream tasks. This allows the model to adapt its reasoning capabilities while preserving the structural understanding gained in Stage 1.

What experiments were performed?

The authors evaluated InstructMol across three distinct categories of drug discovery tasks, comparing it against generalist LLMs (Vicuna, LLaMA, Galactica) and specialist models (ChemBERTa, MolT5):

1. Property Prediction:
   • Regression: Predicting quantum mechanical properties (HOMO, LUMO, Gap) using the QM9 dataset.
   • Classification: Predicting biological activity (BACE, BBBP, HIV) using MoleculeNet.
2. Molecule Description Generation: Generating natural language descriptions of molecules using the ChEBI-20 dataset.
3. Chemical Reaction Analysis:
   • Forward Reaction Prediction: Predicting products from reactants.
   • Reagent Prediction: Identifying necessary reagents.
   • Retrosynthesis: Suggesting reactants for a given product.

Ablation Studies tested the impact of the projector type (Linear vs. MLP), LLM scale (7B vs 13B), and the necessity of the two-stage training approach.

What were the outcomes and conclusions drawn?

• Superiority over Generalists: InstructMol significantly outperformed generalist LLMs (like LLaMA and Galactica) on all tasks, demonstrating the value of incorporating explicit graph modalities.
• Competitive with Specialists: While it did not fully beat highly optimized specialist models (like Uni-Mol) on all classification benchmarks, it significantly reduced the performance gap while retaining the versatility of an LLM.
• Importance of Alignment: Ablation studies confirmed that skipping Stage 1 (Alignment Pre-training) degraded performance, proving that a dedicated phase for projecting graph features into text space is crucial.
• Limitation: The model struggles with highly imbalanced datasets (e.g., HIV) and complex reaction mixtures where mapping multiple graph tokens to text becomes ambiguous.

Reproducibility Details

Data

The training pipeline utilizes distinct datasets for the two stages.

Algorithms

• Two-Stage Training:
  1. Alignment Pre-training: Updates only the Projector. Objective is maximizing probability of generating the target text description given the molecule graph.
  2. Instruction Tuning: Updates Projector + LLM (via LoRA). Objective is auto-regressive language modeling on task-specific instructions.
• LoRA (Low-Rank Adaptation): Applied to the LLM in Stage 2. Rank $r=64$, Scaling $\alpha=16$.
• Optimization: AdamW optimizer. Learning rate starts at 2e-3 (Stage 1) and 8e-5 (Stage 2) with cosine decay. Warm-up ratio 0.03.

Models

• Graph Encoder ($f_g$):
  • Architecture: Graph Isomorphism Network (GIN) with 5 layers.
  • Hidden Dimension: 300.
  • Initialization: MoleculeSTM checkpoint (pre-trained via contrastive learning).
  • Status: Frozen during Stage 2.
• LLM:
  • Base: Vicuna-v1.3-7B.
  • Status: Frozen in Stage 1; LoRA fine-tuned in Stage 2.
• Projector:
  • Architecture: Linear Layer.
  • Function: Maps node-level graph representation $Z_G \in \mathbb{R}^{N \times d}$ to the LLM’s word embedding space dimensions.

Evaluation

• Metric Libraries: RDKit for validity/fingerprints, standard NLP libraries for BLEU/ROUGE.
• Reaction Metrics: Fingerprint Tanimoto Similarity (FTS), Exact Match, Levenshtein distance, and validity (via RDKit).
• Description Metrics: BLEU-2, BLEU-4, ROUGE-1, ROUGE-2, ROUGE-L, METEOR.

Hardware

• Compute: 4 × NVIDIA RTX A6000 (48GB VRAM).
• Training Time:
  • Stage 1: 5 epochs.
  • Stage 2: 20-50 epochs (Description Generation), 10 epochs (Properties/Reactions).
• Batch Size: 128 for both stages.

Citation

@inproceedings{caoInstructMolMultiModalIntegration2025,
  title = {{{InstructMol}}: {{Multi-Modal Integration}} for {{Building}} a {{Versatile}} and {{Reliable Molecular Assistant}} in {{Drug Discovery}}},
  shorttitle = {{{InstructMol}}},
  booktitle = {Proceedings of the 31st {{International Conference}} on {{Computational Linguistics}}},
  author = {Cao, He and Liu, Zijing and Lu, Xingyu and Yao, Yuan and Li, Yu},
  editor = {Rambow, Owen and Wanner, Leo and Apidianaki, Marianna and {Al-Khalifa}, Hend and Eugenio, Barbara Di and Schockaert, Steven},
  year = 2025,
  month = jan,
  pages = {354--379},
  publisher = {Association for Computational Linguistics},
  address = {Abu Dhabi, UAE},
  abstract = {The rapid evolution of artificial intelligence in drug discovery encounters challenges with generalization and extensive training, yet Large Language Models (LLMs) offer promise in reshaping interactions with complex molecular data. Our novel contribution, InstructMol, a multi-modal LLM, effectively aligns molecular structures with natural language via an instruction-tuning approach, utilizing a two-stage training strategy that adeptly combines limited domain-specific data with molecular and textual information. InstructMol showcases substantial performance improvements in drug discovery-related molecular tasks, surpassing leading LLMs and significantly reducing the gap with specialists, thereby establishing a robust foundation for a versatile and dependable drug discovery assistant.}
}