ProCS15: A DFT-based chemical shift predictor for backbone and C β atoms in proteins

Abstract

We present ProCS15: A program that computes the isotropic chemical shielding values of backbone and Cβ atoms given a protein structure in less than a second. ProCS15 is based on around 2.35 million OPBE/6-31G(d,p)//PM6 calculations on tripeptides and small structural models of hydrogen-bonding. The ProCS15-predicted chemical shielding values are compared to experimentally measured chemical shifts for Ubiquitin and the third IgG-binding domain of Protein G through linear regression and yield RMSD values below 2.2, 0.7, and 4.8 ppm for carbon, hydrogen, and nitrogen atoms respectively. These RMSD values are very similar to corresponding RMSD values computed using OPBE/6-31G(d,p) for the entire structure for each protein. The maximum RMSD values can be reduced by using NMR-derived structural ensembles of Ubiquitin. For example, for the largest ensemble the largest RMSD values are 1.7, 0.5, and 3.5 ppm for carbon, hydrogen, and nitrogen. The corresponding RMSD values predicted by several empirical chemical shift predictors range between 0.7 -1.1, 0.2 -0.4, and 1.8 -2.8 ppm for carbon, hydrogen, and nitrogen atoms, respectively.

Untitled

“…Chemical shift prediction using CHARMM minimized X-ray structures The first row of Table 1 lists average RMSD and r values relative to experiment for chemical shift values computed using ProCS15 for 17 protein structures minimized using the CHARMM/CMAP force field. The RMSD values are generally very similar to those computed previously for Ubiquitin and Protein GB3 (Larsen et al, 2015), with the exception of N, where the average RMSD is 0.5 to 0.8 ppm lower. The r values for CA and CB are also quite similar to those obtained previously, but are significantly lower for the remaining nuclei.…”

“…The r values for CA and CB are also quite similar to those obtained previously, but are significantly lower for the remaining nuclei. As we observed previously (Larsen et al, 2015) the RMSD values for ProCS15 are significantly higher (0.5 -1.4 ppm for carbon and N atoms) than those for commonly used empirical chemical shift predictors and very similar to CheShift values, while the corresponding ProCS15 r values are lower than for the empirical methods and similar to CheShift. We now show that the agreement with experiment can be significantly improved for ProCS15 by making relative small changes to the protein structure.…”

“…Analysis of the chemical shift contributions considered by ProCS15 shows that the chemical shift change is due to changes in the f /y and side-chain dihedral angles (s BB cf. Larsen et al (2015)). Comparison of the CHARMM, annealed CHARMM, and annealed ensemble structure, where the CA and HA errors are 0.6 and 0.16 ppm, indicates that the most likely cause is a 22 change in the f angle that alters the interaction with the backbone carbonyl group of Glu136 (Figure 2a).…”

“…In this study we use ProCS15 (Larsen et al, 2015) to predict the corresponding isotropic chemical shielding value s i , which we relate to d i j by d pred,i j = a j s i j + b j…”

“…CamShift (Kohlhoff et al, 2009), PPM One (Li and Brüschweiler, 2015), Sparta+ (Shen and Bax, 2010), shAIC (Nielsen et al, 2012), and ShiftX2 (Han et al, 2011), are typically based on approximate physical models with adjustable parameters that are optimized by minimizing the discrepancy between experimental and predicted chemical shifts computed using protein structures derived from X-ray crystallography. Alternatively, protein chemical shifts can be predicted using computational quantum mechanics (QM), either indirectly using QM-derived models such as SHIFTS Case, 2001, 2002), CheShift (Vila et al, 2009;Martin et al, 2013), and ProCS Larsen et al, 2015), or directly using QM/MM or linear scaling approaches (Johnson and DiLabio, 2009;Zhu et al, 2014;Exner et al, 2012;Sumowski et al, 2014).…”