DP4+App: Finding the Best Balance between Computational Cost and Predictive Capacity in the Structure Elucidation Process by DP4+. Factors Analysis and Automation

FRANCO, B. A.; LUCIANO, E. R.; SAROTTI, A.M.*; ZANARDI, M. M.*

JOURNAL OF NATURAL PRODUCTS

AMER CHEMICAL SOC

Lugar: Washington; Año: 2023 vol. 86 p. 2360 - 2367

0163-3864

DP4+ is one of the most popular methods for the structure elucidation of natural products using NMR calculations. While the method is simple and easy to implement, it requires a series of procedures that can be tedious, coupled with the fact thatits computational demand can be high in certain cases. In this work, we made a substantial improvement to these limitations. First, we deeply explored the effect of molecular mechanics architecture on the DP4+ formalism (MM-DP4+). In addition, a Python applet (DP4+App) was developed to automate the entire process, requiring only the Gaussian NMR output files and a spreadsheet containing the experimental NMR data and labels. The script is designed to use the statistical parameters from the original 24 levels of theory (employing B3LYP/6-31G* geometries) and the new 36 levels explored in this work (over MMFF geometries). Furthermore, it enables the development of customizable methods using any desired level of theory, allowing for a free choice of testmolecules