A foundational review surveying how deep generative models (VAEs, GANs, reinforcement learning) enable inverse molecular design, covering molecular representations, chemical space navigation, and applications from drug discovery to materials engineering.
Lingo3DMol introduces FSMILES, a fragment-based SMILES representation with local and global coordinates, to generate drug-like 3D molecules in protein pockets via a transformer language model.
Proposes Captions as Representations (CaR), where ChatGPT generates textual explanations for SMILES strings that are then used to fine-tune small language models for molecular property prediction.
Demonstrates that standard transformer language models, trained with next-token prediction on sequences from XYZ, CIF, and PDB files, can generate valid 3D molecules, crystals, and protein binding sites competitive with domain-specific 3D generative models.
MolBERT pre-trains a BERT model on SMILES strings using masked language modeling, SMILES equivalence, and physicochemical property prediction as auxiliary tasks, achieving state-of-the-art results on virtual screening and QSAR benchmarks.
MolPMoFiT applies ULMFiT-style transfer learning to QSAR modeling, pre-training an AWD-LSTM on one million ChEMBL molecules and fine-tuning for property prediction on small datasets.
nach0 unifies natural language and SMILES-based chemical tasks in a single encoder-decoder model, achieving competitive results across molecular property prediction, reaction prediction, molecular generation, and biomedical NLP benchmarks.
PASITHEA adapts deep dreaming from computer vision to molecular design, directly optimizing SELFIES-encoded molecules for target chemical properties via gradient-based inversion of a trained regression network.
An extensive benchmark showing that training RNN generative models with randomized (non-canonical) SMILES strings yields more uniform, complete, and closed molecular output domains than canonical SMILES.
Compares RNN-based and Transformer-based chemical language models across three molecular generation tasks of increasing complexity, finding that RNNs excel at local features while Transformers handle large molecules better.
This paper introduces structured state space sequence (S4) models to chemical language modeling, showing they combine the strengths of LSTMs (efficient recurrent generation) and GPTs (holistic sequence learning) for de novo molecular design.
A GRU-based sequence-to-sequence model that learns fixed-length molecular fingerprints by translating SMILES strings to themselves, enabling unsupervised representation learning for drug discovery tasks.