QDS_extract (170114)

QDS_extract <template mol> <template residue name> <qds mol> <qds residue type>

This command allows one to convert the coordinates of a residue in an existing molecule in the O database into a QDS molecule. This new QDS molecule can be used to generate stereo-chemistry data for the residue via the relevant QDS-related commands to assist in rebuilding. Since the new QDS molecule has all the necessary underlying data structures, the new molecule can be extended using the QDS system of commands.

In the following example, the coordinates of Plasmodium falciparum 1-deoxy-D-xylulose-5-phosphate reductoisomerase (PfDXR) in complex with the fosmidomycin analogue FR-900098 (PDB code 3AUA, Umeda, T., Tanaka, N., Kusakabe, Y., Nakanishi, M., Kitade, Y. & Nakamura, K. T. (2011). Sci. Rep. 1, 9.) are used as the template molecule to generate an analogue that will be modified at the alpha-carbon atom of the inhibitor. 

Here is the active site of the PfDXR A-chain in O rainbow colouring, showing the inhibitor, NADPH co-factor, active site ion and the active-site flap running diagonally above the FR-900098:


First we extract a copy of the FR-900098 into the QDS system.

 O > qds_extr

 New> Molecule name of existing template([] to exit):3aua

 New> Template residue to generate a new QDS([] to exit):a701

 New> QDS molecule name ([] to exit):f99

 New> Residue type [QDS]:f99

 New>  Database compressed.

 New>  Atom pre-assigned C1    C_DOUB

 New>  Atom type assignment C1     to C_DOUB

 New>  Atom pre-assigned N1    N_DOUB

 New>  Atom type assignment N1     to N_DOUB

 New>  Atom type assignment O1     to O_DOUB

 New>  Atom type assignment C2     to C_SING

 New>  Atom type assignment O2     to O_SING

 New>  Atom type assignment C3     to C_AROM

 New>  Atom type assignment C4     to C_SING

 New>  Atom type assignment C11    to C_SING

 New>  Atom type assignment PA1    to P_PYRA

 New>  Atom type assignment OP1    to O_DOUB

 New>  Atom type assignment OP2    to O_DOUB

 New>  Atom type assignment OP3    to O_DOUB

The extraction process requires that each atom in the template residue can be identified and assigned to a QDS atom type. The bond lengths and angles of the template residue are taken rather than the 

values defined in the QDS dictionary. A new molecule is generated that can be modified with QDS, as well as an object of the residue; these are called F99 in the above example and the coordinates will look identical to the starting residue in the template at this stage.


Now, from the QDS master menu, I regularise the new residue to confirm all is well (pick ‘Regularise & colour’ from the menu), and decide to introduce a hydroxyl group at the alpha-carbon atom. The latter is achieved by first choosing the type of atom to add (‘O single bond’) and then indicating where to add this atom (‘Extend from ID’, then identifying the atom). In the PDB file this template atom was named C4 and has the same name in the new QDS molecule.

 Heap> read OK

 Refi >  Refining zone Q1           to Q1           in molecule F99

 Refi > Bad outliers at start of regularization

 Refi >

 Refi >  Rmsd Shift    Bonds     Angles  Fixed Torsion

 Refi >       0.00     0.000      0.00      1.55

 Refi >       0.01     0.001      0.04      0.23

 Refi >

 Refi > Bad outliers at end of regularization

 Refi > Accept coordinates?

 New> This action can be undone.

 New> C_SING angle needed

 New> More than one branch is possible

 New>  Option     1C_SING O_SING   120.000

 New>  Option     2C_SING O_SING  -120.000

 New>  Which one [1] :

 New> This action can be undone.

However, because the C4 atom is of type C_SING, two different conformations are possible and one has to be selected by the user.


Next, I add the other possible conformation at C4.


 New> C_SING angle needed

 New> More than one branch is possible

 New>  Option     1C_SING O_SING   120.000

 New>  Option     2C_SING O_SING  -120.000

 New>  Which one [1] :2


We can now evaluate any contacts with the enzyme, for example, and we see that both conformations make reasonable contacts with the side-chain of Ser306 in the enzyme.


 If we had a complex to refine, we could use the QDS tools to make a suitable dictionary for this new residue that we have created.

In target-based drug design, one often synthesises a large number of variants generated from a common core molecule. Once one has created suitable dictionaries and coordinates for the core molecule (with QDS or other software), it becomes straight forward to then use QDS_extract to generate dictionaries and coordinates for the various derivatives.