Evaluating AM1/d-CB1 for chemical glycobiology QM/MM simulations

Krishna K. Govender, Kevin J. Naidoo

Research output: Contribution to journalArticlepeer-review

15 Citations (Scopus)


The newly parametrized AM1/d-CB1 is evaluated for its performance in modeling monosaccharide structure, carbohydrate ring pucker, amino acid proton transfer, DNA base pair interactions, carbohydrate-aromatic η interactions, and phosphates that are prominent in glycosyltransferases. The accuracy of the method in these computations is compared to a comprehensive range of NDDO methods commonly used to study glycan structure and reactivity in chemical biology. AM1/d-CB1 shows significant improvement over existing NDDO type methods in the computation of five and six membered carbohydrate ring pucker free energy surfaces. Moreover, the computation of carbohydrate amino acid interactions commonly present in catalytic domains and binding sites are improved over existing NDDO methods. AM1/d-CB1 shows slight improvement for carbohydrate-aromatic η interactions compared to a commonly used NDDO method (AM1). The method is applied to a glycosyltransferase reaction, where it is the only NDDO method able to achieve an optimized reaction profile. Moreover, a comparison of the geometry optimized computations of the reaction scheme give a transition state energy barrier that best compares with DFT (MPW1K). Overall, AM1/d-CB1 is shown to significantly improve on existing NDDO methods in modeling chemical glycobiological events.

Original languageEnglish
Pages (from-to)4708-4717
Number of pages10
JournalJournal of Chemical Theory and Computation
Issue number10
Publication statusPublished - 14 Oct 2014
Externally publishedYes

ASJC Scopus subject areas

  • Computer Science Applications
  • Physical and Theoretical Chemistry


Dive into the research topics of 'Evaluating AM1/d-CB1 for chemical glycobiology QM/MM simulations'. Together they form a unique fingerprint.

Cite this