State-of-the-art augmented NLP transformer models for direct and single-step retrosynthesis
| Autor: | Guillaume Godin, Pavel Karpov, Ruud van Deursen, Igor V. Tetko |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
FOS: Computer and information sciences
Computer Science - Machine Learning Computational chemistry Computer science Science Reaction mechanisms General Physics and Astronomy Single step Image processing Machine Learning (stat.ML) 010402 general chemistry computer.software_genre 01 natural sciences General Biochemistry Genetics and Molecular Biology Article Machine Learning (cs.LG) Statistics - Machine Learning lcsh:Science Retrosynthetic analysis Transformer (machine learning model) Multidisciplinary Artificial neural network 010405 organic chemistry business.industry Cheminformatics General Chemistry 0104 chemical sciences 3. Good health Test set Beam search lcsh:Q Artificial intelligence business computer Natural language processing |
| Zdroj: | Nature Communications Nature Communications, Vol 11, Iss 1, Pp 1-11 (2020) Nat. Commun. 11:5575 (2020) |
| ISSN: | 2041-1723 |
| Popis: | We investigated the effect of different training scenarios on predicting the (retro)synthesis of chemical compounds using text-like representation of chemical reactions (SMILES) and Natural Language Processing (NLP) neural network Transformer architecture. We showed that data augmentation, which is a powerful method used in image processing, eliminated the effect of data memorization by neural networks and improved their performance for prediction of new sequences. This effect was observed when augmentation was used simultaneously for input and the target data simultaneously. The top-5 accuracy was 84.8% for the prediction of the largest fragment (thus identifying principal transformation for classical retro-synthesis) for the USPTO-50k test dataset, and was achieved by a combination of SMILES augmentation and a beam search algorithm. The same approach provided significantly better results for the prediction of direct reactions from the single-step USPTO-MIT test set. Our model achieved 90.6% top-1 and 96.1% top-5 accuracy for its challenging mixed set and 97% top-5 accuracy for the USPTO-MIT separated set. It also significantly improved results for USPTO-full set single-step retrosynthesis for both top-1 and top-10 accuracies. The appearance frequency of the most abundantly generated SMILES was well correlated with the prediction outcome and can be used as a measure of the quality of reaction prediction. Development of algorithms to predict reactant and reagents given a target molecule is key to accelerate retrosynthesis approaches. Here the authors demonstrate that applying augmentation techniques to the SMILE representation of target data significantly improves the quality of the reaction predictions. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|