Modelación in silico de actividad antimalárica de sustancias orgánicas
Fecha
2017-06
Autores
Roche Llerena, Viviana
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 Licenciatura en Química
Resumen
La malaria es causante de millones de muertes al año y los fármacos utilizados convencionalmente presentan fenómenos de resistencia. Surge así la necesidad de desarrollar nuevos antimaláricos. En este trabajo se realizan regresiones sobre un conjunto químico (Malaria Box) para desarrollar relaciones lineales y no lineales entre las estructuras moleculares y sus actividades correspondientes. En este estudio se modeló la actividad antimalárica (expresada como el EC50) de 317 entidades químicas, frente al parásito Plasmodium falciparum. Las estructuras químicas fueron codificadas mediante índices de derivada de grafos moleculares (GDIs). Estos fueron calculados con el módulo DIVATI del software TOMOCOMD-CARDD. Posteriormente, para seleccionar descriptores con mayor variabilidad, basado en el cálculo de la entropía de Shannon, fue utilizado el software IMMAN. Los modelos de RLM fueron desarrollados con el software MobyDigs. Los modelos de regresión no lineal fueron desarrollados con el software WEKA, mediante Máquinas de Vectores de Soporte. Los resultados de esta modelación muestran superioridad estadística del modelo no lineal con respecto al lineal, con las mismas variables, lo cual sienta las bases para predecir la actividad antimalárica con el uso de GDIs y la técnica de regresión no lineal, para posteriores estudios de cribado virtual y propuesta de potenciales antimaláricos
Malaria causes millions of deaths every year and conventionally used drugs are generating resistance phenomena. Therefore it is necessary to develop new antimalarial drugs urgently. In this work, regression models are developed on a chemical set (Malaria Box) to find linear and nonlinear relations between molecular structures and their corresponding activities. In this study, the antimalarial activity (expressed as the EC50) of 317 chemical entities was modeled against the parasite Plasmodium falciparum. The chemical structures were encoded using discrete molecular graph derivative indices (GDIs). These were calculated using the DIVATI module of the TOMOCOMD-CARDD software. Later, to select descriptors with greater variability, based on the calculation of the Shannon‟s entropy, the IMMAN software was used. The RLM models were found with the MobyDigs software. Nonlinear regression models were found with the WEKA software, using Support Vector Machines. The results of this modeling suggest statistical superiority of the nonlinear model with respect to the linear model, using the same variables, which establish the bases to predict the antimalarial activity using the GDIs and the nonlinear regression technique, for further studies of virtual screening and proposal of antimalarial candidates.
Malaria causes millions of deaths every year and conventionally used drugs are generating resistance phenomena. Therefore it is necessary to develop new antimalarial drugs urgently. In this work, regression models are developed on a chemical set (Malaria Box) to find linear and nonlinear relations between molecular structures and their corresponding activities. In this study, the antimalarial activity (expressed as the EC50) of 317 chemical entities was modeled against the parasite Plasmodium falciparum. The chemical structures were encoded using discrete molecular graph derivative indices (GDIs). These were calculated using the DIVATI module of the TOMOCOMD-CARDD software. Later, to select descriptors with greater variability, based on the calculation of the Shannon‟s entropy, the IMMAN software was used. The RLM models were found with the MobyDigs software. Nonlinear regression models were found with the WEKA software, using Support Vector Machines. The results of this modeling suggest statistical superiority of the nonlinear model with respect to the linear model, using the same variables, which establish the bases to predict the antimalarial activity using the GDIs and the nonlinear regression technique, for further studies of virtual screening and proposal of antimalarial candidates.
Descripción
Palabras clave
Malaria, Epidemiología, Plasmodium falciparum, Diseño de Fármacos, Estructuras Moleculares