Towards a quantum chemical software package utilizing transferable fragments as molecular building blocks

G. Náray-Szabó, György Kramer, Péter Nagy, S. Kugler

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

Computer programs have been developed or are under development for the IBM personal computer that enable their users to get information on atomic charges, electrostatic potentials, conformational and other properties of molecular systems containing H, C, N, O, F, Si, P, S, or Cl atoms. The zero‐order wavefunction is constructed of strictly localized molecular orbitals with fixed atomic orbital coefficients. The wave function can be refined by optimizing these coefficients, i.e., considering inductive effects via a coupled set of 2 × 2 secular equations within the CNDO/2 approximation. Delocalization and exchange effects are accounted for by expanding the wavefunction on a basis of the aforementioned strictly localized orbitals, instead of conventional atomic orbitals, and solving the corresponding SCF equations. Our method has been applied to the study of large systems. We calculated the electrostatic field of the complex of β‐trypsin and basic pancreatic trypsin inhibitor and it has been found that strong field regions more or less coincide with hydration sites. A further potential application of protein electrostatic fields is in NMR spectroscopy. We found a linear correlation between CαH or backbone NH proton chemical shifts and the protein field at the site of the corresponding proton. At last, we propose a simple method to mimic the bulk around atomic clusters modeling crystalline and amorphous silicon. Based on this method we found a linear correlation between atomic net charges and bond angle distortions in silicon clusters with 35 atoms.

Original languageEnglish
Pages (from-to)555-561
Number of pages7
JournalJournal of Computational Chemistry
Volume8
Issue number4
DOIs
Publication statusPublished - 1987

Fingerprint

Wave functions
Software Package
Software packages
Building Blocks
Electrostatic Field
Fragment
Protons
Strictly
Electric fields
Charge
Secular Equation
Proteins
Protein
Amorphous Silicon
Atoms
Aprotinin
Hydration
Personal Computer
Chemical shift
Molecular orbitals

ASJC Scopus subject areas

  • Chemistry(all)
  • Computational Mathematics

Cite this

Towards a quantum chemical software package utilizing transferable fragments as molecular building blocks. / Náray-Szabó, G.; Kramer, György; Nagy, Péter; Kugler, S.

In: Journal of Computational Chemistry, Vol. 8, No. 4, 1987, p. 555-561.

Research output: Contribution to journalArticle

@article{3ad92f01781f4ff092946f7c5416b7c8,
title = "Towards a quantum chemical software package utilizing transferable fragments as molecular building blocks",
abstract = "Computer programs have been developed or are under development for the IBM personal computer that enable their users to get information on atomic charges, electrostatic potentials, conformational and other properties of molecular systems containing H, C, N, O, F, Si, P, S, or Cl atoms. The zero‐order wavefunction is constructed of strictly localized molecular orbitals with fixed atomic orbital coefficients. The wave function can be refined by optimizing these coefficients, i.e., considering inductive effects via a coupled set of 2 × 2 secular equations within the CNDO/2 approximation. Delocalization and exchange effects are accounted for by expanding the wavefunction on a basis of the aforementioned strictly localized orbitals, instead of conventional atomic orbitals, and solving the corresponding SCF equations. Our method has been applied to the study of large systems. We calculated the electrostatic field of the complex of β‐trypsin and basic pancreatic trypsin inhibitor and it has been found that strong field regions more or less coincide with hydration sites. A further potential application of protein electrostatic fields is in NMR spectroscopy. We found a linear correlation between CαH or backbone NH proton chemical shifts and the protein field at the site of the corresponding proton. At last, we propose a simple method to mimic the bulk around atomic clusters modeling crystalline and amorphous silicon. Based on this method we found a linear correlation between atomic net charges and bond angle distortions in silicon clusters with 35 atoms.",
author = "G. N{\'a}ray-Szab{\'o} and Gy{\"o}rgy Kramer and P{\'e}ter Nagy and S. Kugler",
year = "1987",
doi = "10.1002/jcc.540080436",
language = "English",
volume = "8",
pages = "555--561",
journal = "Journal of Computational Chemistry",
issn = "0192-8651",
publisher = "John Wiley and Sons Inc.",
number = "4",

}

TY - JOUR

T1 - Towards a quantum chemical software package utilizing transferable fragments as molecular building blocks

AU - Náray-Szabó, G.

AU - Kramer, György

AU - Nagy, Péter

AU - Kugler, S.

PY - 1987

Y1 - 1987

N2 - Computer programs have been developed or are under development for the IBM personal computer that enable their users to get information on atomic charges, electrostatic potentials, conformational and other properties of molecular systems containing H, C, N, O, F, Si, P, S, or Cl atoms. The zero‐order wavefunction is constructed of strictly localized molecular orbitals with fixed atomic orbital coefficients. The wave function can be refined by optimizing these coefficients, i.e., considering inductive effects via a coupled set of 2 × 2 secular equations within the CNDO/2 approximation. Delocalization and exchange effects are accounted for by expanding the wavefunction on a basis of the aforementioned strictly localized orbitals, instead of conventional atomic orbitals, and solving the corresponding SCF equations. Our method has been applied to the study of large systems. We calculated the electrostatic field of the complex of β‐trypsin and basic pancreatic trypsin inhibitor and it has been found that strong field regions more or less coincide with hydration sites. A further potential application of protein electrostatic fields is in NMR spectroscopy. We found a linear correlation between CαH or backbone NH proton chemical shifts and the protein field at the site of the corresponding proton. At last, we propose a simple method to mimic the bulk around atomic clusters modeling crystalline and amorphous silicon. Based on this method we found a linear correlation between atomic net charges and bond angle distortions in silicon clusters with 35 atoms.

AB - Computer programs have been developed or are under development for the IBM personal computer that enable their users to get information on atomic charges, electrostatic potentials, conformational and other properties of molecular systems containing H, C, N, O, F, Si, P, S, or Cl atoms. The zero‐order wavefunction is constructed of strictly localized molecular orbitals with fixed atomic orbital coefficients. The wave function can be refined by optimizing these coefficients, i.e., considering inductive effects via a coupled set of 2 × 2 secular equations within the CNDO/2 approximation. Delocalization and exchange effects are accounted for by expanding the wavefunction on a basis of the aforementioned strictly localized orbitals, instead of conventional atomic orbitals, and solving the corresponding SCF equations. Our method has been applied to the study of large systems. We calculated the electrostatic field of the complex of β‐trypsin and basic pancreatic trypsin inhibitor and it has been found that strong field regions more or less coincide with hydration sites. A further potential application of protein electrostatic fields is in NMR spectroscopy. We found a linear correlation between CαH or backbone NH proton chemical shifts and the protein field at the site of the corresponding proton. At last, we propose a simple method to mimic the bulk around atomic clusters modeling crystalline and amorphous silicon. Based on this method we found a linear correlation between atomic net charges and bond angle distortions in silicon clusters with 35 atoms.

UR - http://www.scopus.com/inward/record.url?scp=84988122313&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84988122313&partnerID=8YFLogxK

U2 - 10.1002/jcc.540080436

DO - 10.1002/jcc.540080436

M3 - Article

VL - 8

SP - 555

EP - 561

JO - Journal of Computational Chemistry

JF - Journal of Computational Chemistry

SN - 0192-8651

IS - 4

ER -