Efficiency of the multicanonical simulation method as applied to peptides of increasing size: The heptapeptide deltorphin

✍ Scribed by Fatih Yaşar; Handan Arkin; Tarik Çelik; Bernd A. Berg; Hagai Meirovitch

• Publisher: John Wiley and Sons
• Year: 2002
• Tongue: English
• Weight: 191 KB
• Volume: 23
• Category: Article
• ISSN: 0192-8651
• DOI: 10.1002/jcc.10113

✦ Synopsis

Abstract

The advantage of the multicanonical (MUCA) simulation method of Berg and coworkers over the conventional Metropolis method is in its ability to move a system effectively across energy barriers thereby providing results for a wide range of temperatures. However, a MUCA simulation is based on weights (related to the density of states) that should be determined prior to a production run and their calculation is not straightforward. To overcome this difficulty a procedure has been developed by Berg that calculates the MUCA weights automatically. In a previous article (Yaşar et al. J Comput Chem 2000, 14, 1251–1261) we extended this procedure to continuous systems and applied it successfully to the small pentapeptide Leu‐enkephalin. To investigate the performance of the automated MUCA procedure for larger peptides, we apply it here to deltorphin, a linear heptapeptide with bulky side chains (H‐Tyr^1^‐D‐Met^2^‐Phe^3^‐His^4^‐Leu^5^‐Met^6^‐Asp^7^‐NH~2~). As for Leu‐enkephalin, deltorphin is modeled in vacuum by the potential energy function ECEPP. MUCA is found to perform well. A weak second peak is seen for the specific heat, which is given a special attention. By minimizing the energy of structures along the trajectory it is found that MUCA provides a good conformational coverage of the low energy region of the molecule. These latter results are compared with conformational coverage obtained by the Monte Carlo minimization method of Li and Scheraga. © 2002 Wiley Periodicals, Inc. J Comput Chem 23: 1127–1134, 2002