PMFLib – A TOOLKIT FOR FREE ENERGY CALCULATIONS
P. Kulhánek,1,2 L. Mones,3 Z. Střelcová,2 I. Simon,3 M. Fuxreiter,3 and J. Koča1,2
1Central European Institute of Technology, Masaryk University, Kamenice 5, 625 00 Brno, Czech Republic
2National Centre for Biomolecular Research, Faculty of Science, Masaryk University, Kotlářská 2, 611 37 Brno, Czech Republic
3Institute of Enzymology, Biological Research Center, Hungarian Academy of Sciences, Karolina út 29, H-1113 Budapest, Hungary
The majority of chemical and biochemical processes can be described by underlying free energy behavior. This is the case, for instance, in assembly of supramolecular complexes, chemical reactivity, protein folding and protein-ligand binding. It is evident that the calculation of free energies by computer simulations can be very helpful in better understanding of these processes. One of the most commonly used simulation technique providing the free energy is the calculation of potential of mean force (PMF). PMF represents the free energy along a prescribed reaction coordinate and thus provides information about kinetics and thermodynamics of the studied (bio)chemical events.
We have implemented five well established PMF methods into a single suite of programs, which we call PMFLib. Implemented methods are as follows: adaptive biasing force (ABF) , constrained dynamics (CD) , metadynamics (MTD) , umbrella sampling (US) , and string method (SM) . The PMFLib is a set of utilities and libraries written in Fortran90 and C/C++ programming languages providing PMF calculations in transparent and uniform way. PMFLib can be easily plugged into any existing molecular dynamics code and thus to extend its features towards reliable free energy calculations. Currently, interfaces to AMBER, CPMD, and XdynBP molecular dynamics codes are available. PMFLib also offers multiple-walkers (MW)  extension of MTD and ABF methods. Another implemented method improving sampling is replica-exchange molecular dynamics (REMD) . Both MW and REMD extensions are implemented as weakly bound client-server. This allows utilization of PMFLib in calculations employing massive parallel simulations in heterogeneous computer environments, for example in computational grids.
This work has been supported by the Ministry of Education of the Czech Republic (MSM0021622413 and LC06030 to J.K.), the Grant Agency of Czech Republic (GD301/09/H004 to Z.S.), and the European Community's Seventh Framework Program (grant agreement n° 205872 to P.K.).
 E. Darve, D. Rodriguez-Gomez, A. Pohorille, J. Chem. Phys., 128, (2008), 144120.
 J. Schlitter, M. Klahn, Molecular Physics, 101, (2003), 3439-3443.
 A. Laio, M. Parrinello, P. Natl. Acad. Sci. USA, 99, (2002), 12562-12566.
 G. Torrie, J. Valleau, J. Comput. Phys., 23, (1977), 187-199.
 W. E, W. Ren, E. Vanden-Eijnden, Phys. Rev. B, 66, (2002), 052301.
 P. Raiteri, A. Laio, F. Gervasio, C. Micheletti, M. Parrinello, J. Phys. Chem. B, 110, (2006), 3533-3539.
 Y. Sugita, Y. Okamoto, Chem. Phys. Lett., 314, (1999), 141-151.