Efficient model chemistries for peptides. I. General framework and a study of the heterolevel approximation in RHF and MP2 with Pople split‐valence basis sets

Abstract

We present an exhaustive study of more than 250 ab initio potential energy surfaces (PESs) of the model dipeptide HCO-L-Ala-NH2. The model chemistries (MCs) used are constructed as homo-and heterolevels involving possibly different RHF and MP2 calculations for the geometry and the energy. The basis sets used belong to a sample of 39 selected representants from Pople's split-valence families, ranging from the small 3-21G to the large 6-311++G(2df,2pd). The reference PES to which the rest are compared is the MP2/6-311++G(2df,2pd) homolevel, which, as far as we are aware, is the more accurate PES of a dipeptide in the literature. The aim of the study presented is twofold: On the one hand, the evaluation of the influence of polarization and diffuse functions in the basis set, distinguishing between those placed at 1st-row atoms and those placed at hydrogens, as well as the effect of different contraction and valence splitting schemes. On the other hand, the investigation of the heterolevel assumption, which is defined here to be that which states that heterolevel MCs are more efficient than homolevel MCs. The heterolevel approximation is very commonly used in the literature, but it is seldom checked. As far as we know, the only tests for peptides or related systems, have been performed using a small number of conformers, and this is the first time that this potentially very economical approximation is tested in full PESs. In order to achieve these goals, all data sets have been compared and analyzed in a way which captures the nearness concept in the space of MCs. * Corresponding author. E-mail address: pnique@unizar.es 1The most important results of the study are the following: First, that the convergence in method is not achieved in the RHF → MP2 step. Second that the transferability of basis set accuracy from RHF to MP2 is imperfect. These two conclusions lead us to discourage the use of RHF MCs for peptides. Regarding the relative efficiency of the Pople's basis sets, we recommend the inclusion of polarization functions in 1st-row atoms and we discourage the use of basis sets containing doubly-split polarization shells and no diffuse functions. Also, we have found that 6-31G(d) is very efficient for calculating the geometry, and that both the RHF and MP2 infinite basis set limits are approximately reached at 6-311++G(2df,2pd). Finally, related to the heterolevel approximation, we conclude that it is essentially correct for the description of the conformational behaviour of HCO-L-Ala-NH2 both at RHF and MP2. Nevertheless, we place a cautionary remark on the use of RHF geometries with MP2 single-points: Whereas this practice could be accurate enough for short peptides, the accumulation of errors may render it unreliable for longer chains and require the use of MP2 geometries.

The SAAP force field: Development of the single amino acid potentials for 20 proteinogenic amino acids and Monte Carlo molecular simulation for short peptides

“…Accuracy of the HF level was recently claimed to be not sufficient for analyzing potential energy surfaces of the amino acid units in vacuo. 43,44 Despite the suggestions, we chose this calculation level at this moment to obtain the potential energy surfaces as accurately as possible at the cost of a reasonable computation time. It should also be stressed that the effects of the calculation level on the profiles of the potential energy surfaces are much smaller than the solvent effects: the potential surfaces changed more remarkably by changing the solvent than those observed by elevating the calculation level.…”

Influence of ionization on the conformational preferences of peptide models. Ramachandran surfaces of N‐formyl‐glycine amide and N‐formyl‐alanine amide radical cations

“…In the last two decades, many Ramachandran maps have been reported for peptide models of glycine and alanine using molecular mechanics, semiempirical methods and ab initio and DFT calculations 2–24. Other studies have analyzed the interplay between the intrinsic backbone properties and the side chain backbone interactions 25–27.…”

The Quixote project: Collaborative and Open Quantum Chemistry data management in the Internet age

“…These calculations represent an exhaustive study (whose results and aims have been discussed elsewhere [14]), of more than 250 ab initio potential energy surfaces (PESs) of the model dipeptide HCO-L-Ala-NH 2 . The model chemistries investigated are constructed as homo- and heterolevels involving possibly different RHF and MP2 calculations for the geometry and the energy.…”

“…Thus increasing the expense and "accuracy" of a calculation may not always converge to the "correct" solution. As a consequence, the quality of the results does not steadily grow with the computational effort invested, but rather there exist certain tradeoffs that render the relation between them more involved [14-16]. Hence, not only the choice of the more efficient QC method for a given problem among the already existing ones, but also the design of novel model chemistries becomes 'more an art than a science' [17], based more on know-how and empiricism than in a set of systematic procedures.…”