Page 71 out of 225 pages in this book.
Tripos Bookshelf  >  QSAR  >  SAMPLS and Advanced CoMFA Tutorials  >  Database Alignment and Field Classes Tutorial

3.2 Database Alignment and Field Classes Tutorial

• Database Alignment
• Reading in the Biological Data
• Database Alignment
• Final Notes

3.2.1 Database Alignment

You could copy the database from the demo directory and align it afterward, but it is more efficient to align the molecules on a common framework "on the fly" using the database alignment facility.

1. All the ryanoids include a tetrahydropyran which can be used to align them. Retrieve tetrahydropyran from the Fragment Library.

   Build/Edit >>> Get Fragment... >>> Tetrahydropyran

2. Remove all the hydrogens.

   Build/Edit >>> Delete >>> Atom...

   Press Atom Types... in the Atom Expression dialog.

   Check the H check box in the Atom Types dialog and press OK.

   Press OK in the Atom Expression dialog.

3. Save the core structure as a mol2 file so you can retrieve it later. This is a precautionary measure because the contents of this work area will be overwritten during step 4.

   File >>> Save As...

   Enter THP_core in the File field and click on Save.

4. Align all the molecules in the database onto one of them using the molecule on the screen as the basis for the alignment¹.

   File >>> Align Database...

The Database Align Dialog has three fields: one for the database containing the molecules to align, one for the template molecule, and one to indicate the molecule area which contains the substructure common to all molecules. Since only the default work area is currently occupied, that default area is already entered in the appropriate text field. For subsequent discussion, it will be assumed that this is M1.

   Type $TA_DEMO/ryanoids in the Database to Align field. If there is already an entry here, you can select the whole path by double-clicking on it and replace it. Alternatively, you can delete the current entry and leave it empty, or even insert nonsense. Whenever SYBYL cannot interpret the name you supply, it will launch the Database Selection dialog.

   Type ryanodine_1 in the Template Molecule field (the "_1" refers to the structure number given in Welch et al. [Ref. 59]).

   Make sure the New Database and All Molecules radio buttons have been checked and press Go.

   Press OK in the information box warning you that the contents of the molecule areas will be overwritten during the alignment process.

5. Watch the alignment as it proceeds: each molecule in turn is retrieved from the database, brought into a new molecule area and aligned on the template. Once the process is complete, you are prompted for the name of a new database to store the 18 ryanoids in their new orientations.

   Type my_ryanoids at the prompt for the name of the new database and press OK.

6. The new alignments have all been saved, so you can now clear the screen. This will reduce the time SYBYL spends refreshing the screen between commands.

   Build/Edit >>> Zap (Delete) Molecule

   Press All and press OK.

7. Create a spreadsheet from the new database and save it.

   File >>> Molecular Spreadsheet >>> New...

   Select Database and press OK.

   Select my_ryanoids.mdb in the Databases list of the Database Selection dialog and press Open.

   MSS: File >>> Save

   Press OK to accept my_ryanoids.tbl as the file name.

Note About Template Alignment:

Setting the Grid Orientation to Inertial in the Database Align dialog causes the molecule upon which the database is being aligned to be oriented along the coordinate lattice using ORIENT BEST_VIEW (see the Graphics Manual), then frozen (VIEW >>> Freeze View). This makes use of the information embedded in a CoMFA field maximally efficient, will generally improve q², and will make your results reproducible by other workers². We recommend that the template molecule always be oriented in this way before undertaking CoMFA.

            

This feature corresponds to the DATABASE MATCH_ALIGN command.

With Region Focusing, this addresses the concerns raised by Cho and Tropsha [Ref. 20] while avoiding the hazards of a posteriori variable selection. See also Ref. 21.