Loop Modelling 1.0

The database

The LoopFinder uses a database for extraction of essential PDB information regarding the loops. To give LoopFinder some speed, this database holds a small degree of organisation. The first step in this otganisation is that the database is split into three similar parts that are stored in the directories data2, data3, and data4. These three directories hold all data needed for searches with an anchor-length of 2, 3, and 4 Cα atoms, rep

Content per loop

For each loop (with length ranging from 5-26 amino acids) the database holds The coordinates mentioned at the last bullet point are those of the first and last 2, 3, or 4 residues of the loop. The coordinates are stored for 8 Cα atoms; also if the user only wants 2, or 3 anchor residues. In these two cases the last 4, respectively 2 sets of X,Y,Z coordinates are just meaningless random numbers that are not being used by the LoopFinder.

Database file names

In each of the three directories (data2, data3, data4) you find files with names like fort.110, fort.111, etcetera. The numnerical part of these file names are two times the distance between the first and the last Cα in the loop, rounded to the nearest whole number. Given that it seems highly unlikely that anybody will ever search for all loops that match with an allowed error greater that 1.0 Ångstrom, this allows LoopFinder to stop searching after 3 of the 25 database files have been scanned.

Database file format

FORTRAN works much faster with unformatted (only computer readable) files than with formatted (also human readable files). The database therefore consists of unformatted files.
To help programmers work with these files we added the very small FORTRAN program ReadOne.f that reads the smalles database file (data2/fort.105) and prints its contents. You execute and run this program in the main area where you keep all your LoopFinder files with
gfortran -O2 -o ReadOne ReadOne.f
chmod 777 ReadOne

Gert Vriend