MigrationGuide0203

Migration guide from ver. 0.2 to 0.3

The following points are different from ver. 0.2. Version 0.3 is still in development and fast changing. This list is tentative and subject to change.

• ermod now only checks the existence of either "flcuv.tt" in the solution system and "progress.tt" in the reference system. Other files are overwritten without errors. Thus, if users want to redo the calculation, remove only "flcuv.tt" or "progress.tt". , and there are no need to remove other files.

• ermod does not output the self-energy distribution slfeng.XX at default. A key of selfcal is defined to output the self-energy distribution. This new default setup is particular convenient in the calculation of protein solvation. It is now unnecessary for most cases to perform the procedures regarding peread and EcdInfo described in Free energy calculation of protein solvation with explicit solvent.

• gen_structure and gen_input, the semi-automatic generators of the input files for ermod, is available for AMBER, in addition to NAMD and gromacs.

• There was a bug in case that the soln system contains more than one molecules for the solute species. The bug has been fixed, and the ermod procedure is described in Finite concentration of solute: when more than one molecules are present for the solute species. On the user's side, there is nothing special to do for handling the system containing more than one solute molecules.

• When ermod is used with a NAMD trajectory, it is no more necessary to prepare a .xst file. The cell information is directly read from the .dcd file and the file of HISTCELL linked to .xst file is not created by the gen_input script.

• Treatment of asymmetric slab geometry is made possible

• A snapshot can have its own weight; the weight information is saved in a separate file. The name of the file storing the weight is SysWght for the solution and for the solvent, and is SltWght for the isolated solute used for test-particle insertion.

• A key of insstructure is introduced to select some solute structures according to the solute RMSD relative to the reference structure.

• A procedure is defined which conducts a best-fit of the host (presumably protein) to the reference structure and inserts the solute (presumably ligand) within an given range of RMSD.

• The predetermined directories and filenames for slvfe is now flexible.

• The dispersion correction on Lennard-Jones interaction is added in the modified slvfe.F90 program prepared in the extensions directory.