Identificación asistida por computadora de nuevos compuestos líderes con actividad antiinflamatoria
Archivos
Fecha
2009-06-25
Autores
Pedroso Oliva, Yamilka
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Central "Marta Abreu" de las Villas
Resumen
El objetivo fundamental de este estudio fue desarrollar relaciones cuantitativas estructura-actividad (QSAR) para la clasificación y la predicción de la actividad antiinflamatoria, de manera que permita el desarrollo de procesos de tamizaje virtual. Con este fin, una base de datos de 587 compuestos reportados con esta actividad, y utilizando las familias de descriptores lineales y bilineales, estocásticos y no estocásticos, basados en relaciones de enlace implementados en el programa
TOMOCOMD-CARDD, fueron calculados y analizados. Los modelos fueron desarrollados usando el análisis discriminante lineal como técnica estadística de clasificación. Para identificar los descriptores que permitan la separación en dos clases, compuestos con actividad y sin actividad antiinflamatoria fue empleado como método de selección de variables, los análisis de “pasos-haciadelante” y “mejor subconjunto”. Para acceder al poder predictivo de los modelos, se realizó una
validación con una serie de predicción externa de 294 compuestos. El mejor modelo fue sometido además, a un poroceso de validación interna. Los resultados de los análisis indican que los descriptors TOMOCOMD-CARDD totales y locales, proporcionan una adecuada separación de la data en la serie de entrenamiento y en la serie de predicción, respectivamente. Finalmente, todos los modelos se utilizaron para estimar “in silico” la actividad de compuestos con otros usos farmacológicos y nuevas entidades moleculares. Varios fármacos utilizados en la terapéutica actual y nuevos compuesto de síntesis fueron identificados como potenciales compuestos antiinflamatorios.
Se necesita de una posterior corroboración experimental para verificar estos resultados. De forma general podemos concluir que el método TOMOCOMD-CARDD permite el desarrollo de modelos QSAR útiles en el descubrimiento biosílico de nuevos fármacos con actividad anti-inflamatoria.
The aim of this study was to develop a simple quantitative structure-activity relationship (QSAR) for the classification and prediction of anti-inflammatory activity, so as to enable in silico screening. To this end a database of 587 compounds, classified according to whether they had antiinflammatory activity, and using bond-based non-stochastic and stochastic linear and bilinear indices descriptor’s TOMOCOMD-CARDD families were calculated and analyzed. To identify descriptors that allowed separation of the two classes; compounds with and without antiinflammatory activity, analysis of forward stepwise was utilized like variable selection’s method, and models were developed using linear discriminate analysis (LDA). Models predictivity were assessed and validated by the used of an external test set (294 compounds) and method leave-group out, for which predictions were made from the models. The results of the analyses indicated that total and local TOMOCOMD-CARDD descriptors, provided adequate separation of the data in training and test set, respectively. Finally, the models obtained were applied to the virtual screening of chemical compound which allowed the in silico estimation of the activity of compound whit other pharmacological uses as well as new molecular entities. Several drugs current uses in therapy and new series heads was identified as possible anti-inflammatory although the activity of the compounds selected as anti-inflammatory has to be corroborated experimentally. So, the obtained LDA–based QSAR models can be applied to a large set of compounds searching for new candidates as anti-inflammatory. Generally we can conclude that the TOMOCOMD-CARDD descriptors are promising in the development of QSAR models with a view to the biosilic discovery of new drugs with anti-inflammatory activity.
The aim of this study was to develop a simple quantitative structure-activity relationship (QSAR) for the classification and prediction of anti-inflammatory activity, so as to enable in silico screening. To this end a database of 587 compounds, classified according to whether they had antiinflammatory activity, and using bond-based non-stochastic and stochastic linear and bilinear indices descriptor’s TOMOCOMD-CARDD families were calculated and analyzed. To identify descriptors that allowed separation of the two classes; compounds with and without antiinflammatory activity, analysis of forward stepwise was utilized like variable selection’s method, and models were developed using linear discriminate analysis (LDA). Models predictivity were assessed and validated by the used of an external test set (294 compounds) and method leave-group out, for which predictions were made from the models. The results of the analyses indicated that total and local TOMOCOMD-CARDD descriptors, provided adequate separation of the data in training and test set, respectively. Finally, the models obtained were applied to the virtual screening of chemical compound which allowed the in silico estimation of the activity of compound whit other pharmacological uses as well as new molecular entities. Several drugs current uses in therapy and new series heads was identified as possible anti-inflammatory although the activity of the compounds selected as anti-inflammatory has to be corroborated experimentally. So, the obtained LDA–based QSAR models can be applied to a large set of compounds searching for new candidates as anti-inflammatory. Generally we can conclude that the TOMOCOMD-CARDD descriptors are promising in the development of QSAR models with a view to the biosilic discovery of new drugs with anti-inflammatory activity.
Descripción
Palabras clave
TOMOCOMD-CARDD, Descriptores Moleculares, Antiinflamatorios, Anti-Inflammatory Agents