LUMI-lab: a Foundation Model-Driven Autonomous Platform Enabling Discovery of New Ionizable Lipid Designs for mRNA Delivery
Bibliographic record
Abstract
Abstract The complexity of molecular discovery requires autonomous systems that efficiently explore vast and uncharted chemical spaces. While integrating artificial intelligence (AI) with robotic automation has accelerated discovery, its application remains constrained in fields with scarce historical data. One such challenge is the design of lipid nanoparticles (LNPs) for mRNA delivery, which has relied on expert-driven design and is hindered by limited datasets. Here, we introduce LUMI-lab, a self-driving lab (SDL) system that enables efficient learning with minimal wet-lab data by integrating a molecular foundation model with an automated active-learning experimental workflow. Through ten iterative cycles, LUMI-lab synthesized and evaluated over 1,700 LNPs, identifying ionizable lipids with superior mRNA transfection potency in human bronchial cells compared to clinically approved benchmarks. Unexpectedly, it autonomously uncovered brominated lipid tails as a novel feature enhancing mRNA delivery. In vivo validation further confirmed that inhalation of LNPs containing the top-performing lipid, LUMI-6, achieved 20.3% gene editing efficacy in lung epithelial cells in murine models, surpassing the highest efficiency reported for inhaled LNP-mediated CRISPR-Cas9 delivery in mice to our knowledge. These findings demonstrate LUMI-lab as a powerful, data-efficient platform for advancing mRNA delivery, highlighting the potential of AI-driven autonomous systems to accelerate innovation in material science and therapeutic discovery.
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 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".