Nagesh Chakka wrote: > I am trying to use Modeller to obtain a model using the structure 1XU0. > I have created the three files needed for Modeller > In doing so, the final model XlPrP2.B99990004 does not contain all the > atoms for a particular residue. > > For example, the following is the part of the model for one amino acid > > ATOM 1 N ILE 1 14.910 4.779 7.839 1.00 41.08 > 1SG 2 > ATOM 2 CA ILE 1 13.848 4.061 8.581 1.00 41.08 > 1SG 3 > ATOM 3 CB ILE 1 12.918 3.371 7.630 1.00 41.08 > 1SG 4 > ATOM 4 CG2 ILE 1 11.924 2.529 8.447 1.00 41.08 > 1SG 5 > ATOM 5 CG1 ILE 1 12.242 4.416 6.730 1.00 41.08 > 1SG 6 > ATOM 6 CD1 ILE 1 11.473 3.813 5.557 1.00 41.08 > 1SG 7 > ATOM 7 C ILE 1 14.422 3.050 9.510 1.00 41.08 > 1SG 8 > ATOM 8 O ILE 1 15.261 2.235 9.128 1.00 41.08 > 1SG 9 > > Atoms from the structure file > > ATOM 1 N ILE A 125 15.197 -5.772 -0.208 1.00 0.00 > N > ATOM 2 CA ILE A 125 14.432 -5.479 0.993 1.00 0.00 > C > ATOM 3 C ILE A 125 14.890 -6.453 2.086 1.00 0.00 > C > ATOM 4 O ILE A 125 14.874 -7.667 1.884 1.00 0.00 > O > ATOM 5 CB ILE A 125 12.922 -5.586 0.661 1.00 0.00 > C > ATOM 6 CG1 ILE A 125 12.526 -4.523 -0.390 1.00 0.00 > C > ATOM 7 CG2 ILE A 125 12.073 -5.416 1.930 1.00 0.00 > C > ATOM 8 CD1 ILE A 125 11.131 -4.703 -1.000 1.00 0.00 > C > ATOM 9 H ILE A 125 14.783 -6.386 -0.892 1.00 0.00 > H > ATOM 10 HA ILE A 125 14.656 -4.459 1.305 1.00 0.00 > H > ATOM 11 HB ILE A 125 12.723 -6.575 0.246 1.00 0.00 > H > ATOM 12 1HG1 ILE A 125 12.591 -3.543 0.075 1.00 0.00 > H > ATOM 13 2HG1 ILE A 125 13.228 -4.540 -1.223 1.00 0.00 > H > ATOM 14 1HG2 ILE A 125 12.338 -4.494 2.447 1.00 0.00 > H > ATOM 15 2HG2 ILE A 125 11.014 -5.400 1.691 1.00 0.00 > H > ATOM 16 3HG2 ILE A 125 12.237 -6.268 2.586 1.00 0.00 > H > ATOM 17 1HD1 ILE A 125 10.989 -3.972 -1.797 1.00 0.00 > H > ATOM 18 2HD1 ILE A 125 11.032 -5.707 -1.414 1.00 0.00 > H > ATOM 19 3HD1 ILE A 125 10.362 -4.542 -0.250 1.00 0.00 > H > > What is the reason for the loss of the atoms and how can I fix this.
Modeller builds models, by default, using a heavy atom only topology, so there are no hydrogen atoms. (Including hydrogen atoms greatly increases the CPU time needed to build the models, and generally adds very little information, since most templates don't have hydrogen information, so the hydrogen positions are largely determined by geometry.)
You can build all hydrogen models by using a different topology library. For modern versions of Modeller, just use the 'allhmodel' class (http://salilab.org/modeller/manual/node28.html). For old versions using TOP, turn on HYDROGEN_IO to read hydrogens from the template structures, and set TOPLIB to request the all-atom topology. See http://salilab.org/modeller/7v7/FAQ.html#15
Ben Webb, Modeller Caretaker