Randomized SMILES strings improve the quality of molecular generative models

By Josep Arús‐Pous, Simon Johansson, Oleksii Prykhodko, Esben Jannik Bjerrum, Christian Tyrchan, Jean-Louis Reymond, Hongming Chen, Ola Engkvist. 2019

340 citations recorded by the source corpus.

Recurrent Neural Networks (RNNs) trained with a set of molecules represented as unique (canonical) SMILES strings, have shown the capacity to create large chemical spaces of valid and meaningful structures. Herein we perform an extensive benchmark on models trained with subsets of GDB-13 of different sizes (1 million, 10,000 and 1000), with different SMILES variants (canonical, randomized and DeepSMILES), with two different recurrent cell types (LSTM and GRU) and with different hyperparameter combinations. To guide the benchmarks new metrics were developed that define how well a model has generalized the training set. The generated chemical space is evaluated with respect to its uniformity, closedness and completeness. Results show that models that use LSTM cells trained with 1 million randomized SMILES, a non-unique molecular string representation, are able to generalize to larger chemical spaces than the other approaches and they represent more accurately the target chemical space. Specifically, a model was trained with randomized SMILES that was able to generate almost all molecules from GDB-13 with a quasi-uniform probability. Models trained with smaller samples show an even bigger improvement when trained with randomized SMILES models. Additionally, models were trained on molecules obtained from ChEMBL and illustrate again that training with randomized SMILES lead to models having a better representation of the drug-like chemical space. Namely, the model trained with randomized SMILES was able to generate at least double the amount of unique molecules with the same distribution of properties comparing to one trained with canonical SMILES.

Subfields: Computational Theory and Mathematics, Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

In Molecular Biology

Citations in the corpus

Molecular Sets (MOSES): A Benchmarking Platform for Molecular Generation Models (2018, background): “, 1989), ambiguity of SMILES can also serve as augmentation and improve generative models (Arús-Pous et al., 2019).”
Mol-CycleGAN: a generative model for molecular optimization (2019, background): “[21] show that randomization of SMILES can substantially improve the quality of generated molecules.”
A de novo molecular generation method using latent vector based generative adversarial network (2019, background): “Then, structure generation is basically a sampling process following the learned probability distribution [7, 8, 13, 14].”
Generating 3D Molecular Structures Conditional on a Receptor Binding Site with Deep Generative Models (2020, background): “Additionally, similar molecules can have very different strings due to lack of permutation invariance, and randomization of SMILES traversal order was proposed to improve the quality of generated molecules [26].”
Molecular representations in AI-driven drug discovery: a review and practical guide (2020, methodology): “Additionally, the use of string representations such as SMILES has seen a lot of unexpected success in the field of de novo molecular design using tools from NLP. Data augmentation can be done for many applications using randomized SMILES [51, 138].”