LightChem: A Lightweight Domain-Specific Language Model for Molecular Property and Reaction Prediction in Chemistry
Bibliographic record
Abstract
Large language models (LLMs) have achieved breakthroughs in natural language processing, yet their use in chemistry is limited by insufficient domain knowledge, high computational demands, and poor scalability. To overcome these challenges, we developed LightChem, a lightweight model that integrates retrieval-augmented generation (RAG) and tool-calling capabilities, providing a platform driven by chemical knowledge and theoretical calculations. LightChem adopts a three-layer architecture, data aggregation, intelligent retrieval, and specialized tool integration, supported by the self-developed RAPTOR system for real-time literature updates. The platform incorporates a suite of computational modules, including automated molecular property prediction (e.g., PoLogP), retrosynthetic planning via ReSynZ, and scalable quantum chemistry packages such as CIM and GEBF, enabling high-accuracy simulations of complex systems. Case studies demonstrate successful predictions of lipophilicity, excitation/emission spectra, and synthetic pathways for drug-like molecules, while large-scale calculations reproduce binding energies in zeolite clusters and excitation energies in GFP chromophores with accuracy comparable to experimental values. Beyond core computations, LightChem features a user-friendly web interface that connects sophisticated algorithms with practical laboratory workflows. It provides modules for reagent management, synthesis assessment, safety evaluation, and experimental design, thereby lowering the barrier for non-specialists and supporting systematic research in photosensitive and functional molecules. Applications to gold nanoclusters and zeolites further highlight the platform’s potential in photocatalysis and optical materials, while also revealing current limitations in predicting HOMO–LUMO gaps and optical properties of complex aggregates. LightChem demonstrates good performance in molecular property prediction, retrosynthetic planning, and large-scale quantum simulations, making it a versatile assistant for chemical research, bridging knowledge-driven reasoning with first-principles accuracy.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.004 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.010 | 0.005 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".