Abstract The modern process of new drug discovery and development is an exciting, yet a challenging, endeavor. Although it can result in significant financial income and meet the medical needs of patients, it ultimately may result in failure. To achieve a fast and successful new product discovery and development process, natural products which are evolutionarily optimized as drug-like molecules have gained great attention as better potential sources of new chemical entities. Historically, plant species containing berberine are used in various traditional phytotherapy. However, despite the various therapeutic effects it exerts, berberine is not yet developed into a drug product. Addressing the barriers that hinder its successful development and the efforts made to overcome them is thus crucial. The toxicological and pharmacokinetic properties of berberine are the main barriers towards its development into a marketed drug product. It has low aqueous solubility, poor absorption, fast metabolism, and wide tissue distribution which lead to low bioavailability limiting its clinical application. Synthetic berberine derivatives with improved properties are suggested as better alternatives for further development and future therapeutic application. Hence, this paper summarizes the preclinical research studies conducted in the last decade to reveal the therapeutic potential of synthetic berberine derivatives for the treatment of various diseases and hence achieve successful berberine-based drug development in the future. To exploit the value of natural products as a source of leads for the development of effective drugs, collaboration among the different discovery and development scientists is essential.

Abstract: The traditional extreme learning machine has significant disadvantages, including slow training, difficulty in selecting parameters, and difficulty in setting the singularity and the data sample. A prediction model of an improved Online Sequential Extreme Learning Machine (IOS-ELM) of daily reference crop evapotranspiration is therefore examined in this paper. The different manipulation of the inverse of the matrix is made according to the optimal solution and using a regularization factor at the same time in the model. The flexibility of the IOS-ELM in ET0 modeling was assessed using the original meteorological data (Tmax, Tm, Tmin, n, Uh, RHm, φ, Z) of the years 1971-2014 in Yulin, Ankang, Hanzhong, and Xi’an of Shaanxi, China. Those eight parameters were used as the input, while the reference evapotranspiration values were the output. In addition, the ELM, LSSVM, Hargreaves, Priestley-Taylor, Mc Cloud and IOS-ELM models were tested against the FAO-56 PM model by the performance criteria. The experimental results demonstrate that the performance of IOS-ELM was better than the ELM and LSSVM and significantly better than the other empirical models. Furthermore, when the total ET0 estimation of the models was compared by the relative error, the results of the intelligent algorithms were better than empirical models at rates lower than 5%, but the gross ET0 empirical models mainly had 12% to 64.60% relative error. This research could provide a reference to accurate ET0 estimation by meteorological data and give accurate predictions of crop water requirements, resulting in intelligent irrigation decisions in Shaanxi.

Resumen: La máquina de aprendizaje extremo tradicional tiene desventajas significativas, tales como entrenamiento lento, dificultad en la selección de parámetros y dificultad en establecer la singularidad y la muestra de datos. Por lo tanto, en el presente artículo se examina un modelo de predicción de una máquina de aprendizaje extremo secuencial en línea mejorada (IOS-ELM) de la evapotranspiración de referencia diaria de cultivos. La diferente manipulación de la inversa de la matriz se hace de acuerdo con la solución óptima y utilizando un factor de regularización al mismo tiempo en el modelo. La flexibilidad de la IOS-ELM en la modelación de la ET0 se evaluó empleando los datos meteorológicos originales (Tmax, Tm, Tmin, n, Uh, RHm, φ, Z) de los años 1971-2014 en Yulin, Ankang, Hanzhong, y Xi’an en Shaanxi, China. Estos ocho parámetros se usaron como insumos o datos de entrada, mientras que los valores de la evapotranspiración de referencia fueron los datos de salida o el producto. Asimismo, se probaron los modelos ELM, LSSVM, Hargreaves, Priestley-Taylor, Mc Cloud y IOS-ELM contra el modelo FAO-56 PM mediante los criterios de desempeño. Los resultados experimentales demuestran que el desempeño de IOS-ELM fue mejor que le de ELM y LSSVM y significativamente mejor que los demás modelos empíricos. Más aún, al comparar la estimación total de ET0 de los modelos mediante el error relativo, los resultados de los algoritmos inteligentes fueron mejores que los modelos empíricos a índices inferiores a 5%, pero los modelos empíricos de ET0 bruta tuvieron un error relativo de 12 a 64.60%. Esta investigación podría proporcionar una referencia para la estimación precisa de ET0 mediante datos meteorológicos y proporcionar predicciones precisas de los requerimientos de agua de los cultivos, lo cual resultaría en decisiones de riego inteligentes en Shaanxi.