Identificación in silico de nuevos compuestos inhibidores de Termolisina derivados de la planta Boldoa purpurascens
Fecha
2018
Autores
Ulloa Quintero, Adriana
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Central “Marta Abreu” de Las Villas. Facultad de Química Farmacia. Departamento de Farmacia
Resumen
Esta investigación se centra en la identificación de potenciales fármacos
inhibidores de la Termolisina (TLN), a partir de un grupo de compuestos
aislados de la planta Boldoa purpurascens, que pueden servir como cabezas
de serie para el desarrollo de fármacos antihipertensivos. En este trabajo se
emplean modelos QSAR para la clasificación de compuestos inhibidores de la
TLN. Se utilizó una base de datos de 191 compuestos reportados en la
literatura y fueron calculados los descriptores implementados en el software
DRAGON. Para desarrollar los modelos de clasificación se trabajó con técnicas
de minería de datos implementadas en el software WEKA. Los modelos
obtenidos mostraron valores de precisión superiores al 85 %, y la razón de
falsos positivos se mantuvo, en cada modelo, con valores adecuados, por
debajo de 15. Para la validación de los modelos se llevaron a cabo estrategias
de validación cruzada y validación externa usando una serie de predicción.
Finalmente, los modelos fueron empleado para el cribado virtual de ocho
flavonoides y sus potenciales productos de hidrólisis, total o parcial, para
descubrir nuevos inhibidores de la Termolisina identificándose que la totalidad
de los compuestos inhiben la enzima anteriormente mencionada.
This research focuses on the identification of potential drugs inhibitors of Termolysin (TLN), from a group of compounds isolated from the plant Boldoa purpurascens, which can serve as seeds for the development of antihypertensive drugs. In this work, QSAR models are used for the classification of inhibitory compounds of TLN. A database of 191 compounds reported in the literature was used and the descriptors implemented in the DRAGON software were calculated. In order to develop the classification models, we worked with data mining techniques implemented in the WEKA software. The obtained models showed values of precision superior to 85 %, and the reason of false positives was maintained, in each model, with adequate values, below 15. For the validation of the models, cross-validation and external validation strategies were carried out, using a prediction serie. Finally, the models were used for the virtual screening of eight flavonoids and their potential hydrolysis products, total or partial, to discover new inhibitors of Termolysin, identifying that all the compounds inhibit the aforementioned enzyme.
This research focuses on the identification of potential drugs inhibitors of Termolysin (TLN), from a group of compounds isolated from the plant Boldoa purpurascens, which can serve as seeds for the development of antihypertensive drugs. In this work, QSAR models are used for the classification of inhibitory compounds of TLN. A database of 191 compounds reported in the literature was used and the descriptors implemented in the DRAGON software were calculated. In order to develop the classification models, we worked with data mining techniques implemented in the WEKA software. The obtained models showed values of precision superior to 85 %, and the reason of false positives was maintained, in each model, with adequate values, below 15. For the validation of the models, cross-validation and external validation strategies were carried out, using a prediction serie. Finally, the models were used for the virtual screening of eight flavonoids and their potential hydrolysis products, total or partial, to discover new inhibitors of Termolysin, identifying that all the compounds inhibit the aforementioned enzyme.
Descripción
Palabras clave
Plantas Medicinales, Boldoa Purpurascens, Fármacos Antihipertensivos, Medicina Tradicional, Descriptores Moleculares