Characterising the glass transition temperature-structure relationship through a recurrent neural network
Quantitative structure-property relationship (QSPR) is a powerful analytical method to find correlations between the structure of a molecule and its physicochemical properties. The glass transition temperature (Tg) is one of the most reported properties, and its characterisation is critical for tuni...
| Autores: | , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2023 |
| País: | España |
| Institución: | Consejo Superior de Investigaciones Científicas (CSIC) |
| Repositorio: | DIGITAL.CSIC. Repositorio Institucional del CSIC |
| OAI Identifier: | oai:digital.csic.es:10261/342175 |
| Acceso en línea: | http://hdl.handle.net/10261/342175 |
| Access Level: | acceso abierto |
| Palabra clave: | QSPR Machine learning Molecular glass former Amino acid RNN |
| Sumario: | Quantitative structure-property relationship (QSPR) is a powerful analytical method to find correlations between the structure of a molecule and its physicochemical properties. The glass transition temperature (Tg) is one of the most reported properties, and its characterisation is critical for tuning the physical properties of materials. In this work, we explore the use of machine learning in the field of QSPR by developing a recurrent neural network (RNN) that relates the chemical structure and the glass transition temperature of molecular glass formers. In addition, we performed a chemical embedding from the last hidden layer of the RNN architecture into an m-dimensional Tg-oriented space. Then, we test the model to predict the glass transition temperature of essential amino acids and peptides. The results are very promising and they can open the door for exploring and designing new materials. |
|---|