Biocomputing

Coarse-grained model for peptides

I am working on a new coarse-grained (CG) model for peptides and protein systems compatible with the coarse-grained model for lipids and surfactants recently published by Marrink et al. (J. Phys. Chem. B, 2004, 108, 750-760). A small number of coarse grained atom types are defined, which interact using a few discrete levels of interaction. This model is computationally efficient, with an increase in simulation speed of about 3 orders of magnitude compared to atomistic models, due to the reduction of the number of degrees of freedom and the use of short range potentials.

Despite its simplistic nature, the model proves to be reasonably accurate in the prediction of partitioning and orientation for a series of pentapeptides with sequence Ace-WLXLL in water/alkane systems, as well as for the helical transmembrane peptide WALP23 in a CG lipid bilayer. Orientation and aggregation behavior were also compared for WALP23, alamethicin and poly-leucine peptide dimers in a DOPC bilayer. Comparisons with atomistic simulations and, where possible, with experimental data, are used here to validate the model. At the present stage of development, the potential function used cannot provide predictions on peptide conformation; therefore, the model is not meant to be used for protein folding applications.

A lipid bilayer structure library for molecular simulations

Lipid bilayers are fundamental components of biological membranes. The most biologically relevant state of lipid bilayers is the fully hydrated fluid phase (fluid-disordered Lα or liquid-ordered). Due to their fluidity, it is not possible to obtain experimentally atomic level structures of single bilayers. Computer simulations have been widely used to study lipid bilayers, providing detailed structural information that can aid the interpretation of experimental results. A large body of literature is available on simulations of lipid bilayers, but different studies used different simulation conditions, which causes problems when comparing results for different lipids.

We are developing a library of structures of lipid bilayers, generated using molecular simulations in identical conditions. The library provides comparable structural data for 60 lipid bilayers and equilibrium starting structures for membrane protein simulations. Four head groups and 15 pairs of fatty acid chains were used, including the most common lipid molecules. Important structural parameters, such as the area per lipid and the electron density profiles, can be observed from our simulations on series of related lipids.

Peptide Folding

Folding of helical peptides has been studied for a long time, but not too many experimental data are available both on the thermodynamics and the kinetics of the process. I've been studying RN24, a 13-residue peptide derived from the N-terminus of Ribonuclease A. I'm using microsecond scale molecular dynamics simulations to investigate the free energy landscape of the peptide and the mechanism of helix folding.

The RN24 peptide is partially helical at low temperature (275 K) and low pH. Red dotted lines represent experimentally measured interproton distances (from Osterhout, J. J. J, Baldwin, R. L, York, E. J, Stewart, J. M, Dyson, H. J, & Wright, P. E. (1989) Biochemistry 28, 7059-7064.)

The peptide, which is partially helical according to NMR experiments, explores many different conformations including helices, beta-sheet and beta-hairpin conformations, and folds at least partially into an helix many times during the simulations. Analyzing the hydrogen bonding pattern as a function of simulation time (using a program that I wrote) it is possible to gain lots of informations about helix folding mechanism, both for helix nucleation and propagation.