The Sup command superposes two objects based on the selected atoms. The command extension (Obj,Mol,Res,Atom) thus defines what is used to calculate the superposition matrix. During the actual superposition step, YASARA always transforms the entire object containing the source selection.

When superposing proteins with a different sequence (and hence a different number of atoms), always exclude the side-chains and superpose only CA (SupAtom CA Obj 1,CA Obj 2) or backbone atoms. Also make sure that both proteins contain the same number of residues, otherwise select only a subset (SupAtom CA Res 10-25 Obj 1,CA Res 20-35 Obj 2).

The Sup command does not perform a structural alignment like AlignObj, so you have to define yourself which atoms are to be superposed. If you set match to yes, YASARA will help you a bit by matching the atom/residue names in both selections, keeping only atoms that are present in both.

Be careful about the various definitions for 'Backbone RMSD'. SupAtom Backbone Obj 1,Backbone Obj 2 will use *all* backbone atoms, including the oxygens (possibly two at the C-terminus), and any hydrogens attached to the CA atom and the N-terminal nitrogen. Usual definitions for backbone RMSD ignore the hydrogens (SupAtom Backbone Element !H Obj 1,Backbone Element !H Obj 2), sometimes also the second C-terminal oxygen (SupAtom N CA C O OT1 Obj 1,N CA C O OT1 Obj 2) or even all oxygens (SupAtom N CA C Obj 1,N CA C Obj 2). If accidentally the two structures to superpose differ at the C-terminal end (one has a carboxyl group, the other only one oxygen), use one of the previous two examples if you want to include the terminal oxygen, or 'SupAtom N CA C O Obj 1,N CA C O Obj 2,Match=Yes' if you don't.

The Sup command returns the RMSDs calculated after superposition, according to the formula shown at the RMSD command. Here is an example for calculating a pairwise RMSD matrix:

# CALCULATE A PAIRWISE Calpha RMSD MATRIX
Tabulate SupAtom CA,CA
SaveTab default,RMSDMatrix,(Objects),7.4f,Pairwise RMSD-Matrix

Another problem inherent to superpositions is addressed by the Flip flag: chemical equivalence. A typical example is the phenyl ring in phenylalanine. If you rotate the ring by 180 degrees around the chi-2 dihedral angle, the result is exactly the same, still you will obtain a large RMSD if you try to superpose atoms from opposite sides of the ring. If flip is set to 'yes', YASARA will automatically flip chemically equivalent atoms to minimize the RMSD. The term 'chemically equivalent' includes bond orders , which becomes important in connection with resonance bonds, e.g. in carboxyl groups. If the carboxyl group is neutral, one oxygen makes a double bond, and the other one two single bonds, hence they are not equivalent. If the proton dissociates, both oxygens make one bond of order 1.5 (colored red), they become equivalent and may consequently be flipped during the superposition. Especially when loading small molecules in formats like mol2 that do not support resonance bonds, it is required to assign resonance bonds before doing a superposition with flipping enabled. Also note that flipping chemically equivalent groups may cause deviations from naming conventions . That's why YASARA never flips the two methyl groups of Val and Leu residues, even if that would lower the RMSD.

Finally, the Unit parameter determines the granularity of the returned superposition RMSDs. By default RMSDs are calculated per object, but this can easily be changed to a per-molecule, per-residue or per-atom basis by choosing the appropriate unit.

A hint for superposing structures with a circular permutation : the order in which you select the residues to superpose is meaningless, since it is the order of the residues in the soup that counts. If one of the structures to superpose has a circular permutation, the solution is thus to split the molecule in the middle, swap the two resulting objects, then join them again and finally perform the superposition.

When analyzing a simulation run with periodic boundaries , the superposition on the initial starting structure can give unexpected results if two or more molecules are superposed together: as soon as a molecule crosses a periodic boundary, with two halves at opposite sides of the cell, the atom coordinates are obviously not suited for a superposition. YASARA automatically corrects the coordinates by wraping one half back before performing the superposition. The only indicator that wraping is actually needed is the presence of a chemical bond with an enormous length, that stretches from one side of the cell to the other. Since it is essentially a random decision which of the two halves to wrap, there is no way to guarantee that two separate molecules that are not linked by a bond will end up in a single unique relative position. Especially if the protein is an oligomer, the wraping will not always reproduce the native state and may thus result in arbitrarily high superposition RMSDs with respect to the initial structure. The obvious solution is to perform superposition and RMSD-analysis on a per-molecule basis.

SupObj 1,2

Take object 1 and put in on top of object 2 so that the coordinate RMSD is minimal. Both objects must contain the same number of atoms.

SupObj 1,2,Match=Yes

As above, but use only those atoms that are present in both objects.

SupAtom CA Obj 1,CA Obj 2

Superpose object 1 on object 2 using the CA atoms only.

SupRes SecStr Helix Obj 1,SecStr Helix Obj 2

Superpose object 1 on object 2 using all residues in helical regions. These residues must be the same in both objects.

SupAtom CA SecStr Helix Obj 1,CA SecStr Helix Obj 2

As above, now the residues can differ, but both objects must still have the same number of residues in helical regions.

SupObj 1,3,Flip=Yes

Superpose object 1 on object 3, flipping chemically equivalent atoms to minimize the RMSD.

SupObj 1,5,Unit=Res

Superpose object 1 on object 5 and list the RMSD per residue.

rmsd = SupObj 1,2

Superpose object 1 on 2 and assign the resulting RMSD to variable 'rmsd'.

rmsdlist() = SupObj all,all

Superpose all objects on all objects and assign the RMSD matrix to 'rmsdlist'.

Tabulate SupAtom CA,CA

Superpose all objects on all objects using CA atoms and tabulate the RMSD matrix.

Tabulate SupAtom Backbone,Backbone,Unit=Res

Superpose all objects on all objects using backbone atoms and for each superposition tabulate the RMSDs per residue.