Journal of Computational Chemistry
© Wiley Periodicals, Inc.
Edited By: Charles L. Brooks III, Masahiro Ehara, Gernot Frenking, and Peter R. Schreiner
Impact Factor: 3.648
ISI Journal Citation Reports © Ranking: 2015: 41/163 (Chemistry Multidisciplinary)
Online ISSN: 1096-987X
Recently Published Articles
- “Solvent hydrogen-bond occlusion”: A new model of polar desolvation for biomolecular energetics
Andrea Bazzoli and John Karanicolas
Version of Record online: 20 MAR 2017 | DOI: 10.1002/jcc.24740
The “Solvent Hydrogen bond Occlusion” approach assigns desolvation free energies for individual polar groups, by evaluating the extent to which neighboring atoms prevent the polar group from engaging in hydrogen bonds with solvent. A single probe water molecule is considered, which can occupy grid points around the polar group of interest; the energetics on the grid reflect the preferred hydrogen bonding geometry for the polar atom of interest (color gradient). Neighboring atoms (shown in gray) sterically occlude the probe water from certain locations on the grid: by writing a partition function that sums over these grid points, we can explicitly evaluate the desolvation free energy due to these occluding atoms.
- Study of the cold charge transfer state separation at the TQ1/PC71BM interface
Riccardo Volpi and Mathieu Linares
Version of Record online: 20 MAR 2017 | DOI: 10.1002/jcc.24776
We study the charge transfer (CT) state separation at the interface between TQ1 polymer and PC71BM, two materials used in organic solar cells. With the CT state splitting diagram, we can determine the distribution of rates in function of the electric field for the separation of hole and electron at the interface and for their conduction in the bulk. Kinetic Monte Carlo simulations performed at interesting electric fields allow us to establish relationship between morphology and efficiency of the CT state splitting.
- Numerical interpretation of molecular surface field in dielectric modeling of solvation
Changhao Wang, Li Xiao and Ray Luo
Version of Record online: 20 MAR 2017 | DOI: 10.1002/jcc.24782
A semi-log plot of mean absolute errors of atomic dielectric boundary forces (kcal/mol-e-Å) versus grid spacing (tested from 1/16 Å to 1/2 Å) for a nucleic acid base pair with multiple surface field fitting methods: one-sided first-order (black circle), one-sided second-order (black square), two-sided first-order (red circle), two-sided second-order (red square), and finally, the approximated one-dimensional method (blue star). Our analysis shows that the efficient one-dimensional method achieves similar accuracy as the more expensive second-order method but with a fraction of the computational cost.
- Computational study of the reactivity of cytosine derivatives
Jihène Jerbi and Michael Springborg
Version of Record online: 20 MAR 2017 | DOI: 10.1002/jcc.24781
DNA demethylation can be both passive and active. The passive process is related to a dilution of the 5hmC during cell divisions, whereas the active process involves successive TET-mediated conversions of 5-hmC to 5-formyl-cytosine (5fC) and 5-carboxyl-cytosine (5caC), both of which can be transformed back to the unmodified cytosine through the base excision repair (BER) mechanism. The role of DNA demethylation in the development of cancer has been studied only little in the past. Therefore, a computational study is useful to identify relationships between reactivity and stability for the modified compounds to understand their biological functionalities
- FFLUX: Transferability of polarizable machine-learned electrostatics in peptide chains
Timothy L. Fletcher and Paul L. A. Popelier
Version of Record online: 10 MAR 2017 | DOI: 10.1002/jcc.24775
The machine learning method kriging is trained to predict the net atomic charge of alpha carbons in oligo-peptides. Amino acids can be sorted into “groups” according to their influence on a neighboring alanine residue, leading to kriging models that can predict for an entire group. A dataset using multiple systems for its training data and different systems in the test data is termed a transferred dataset. This is the basis of transferable models in the force field under development called FFLUX. Precursor papers: Fletcher and Popelier, Theor. Chem. Acc. 2015, 134, 135:1 and Fletcher and Popelier, J. Comput. Chem. 2017, 38, 336.