Chapter 38. Output from GRID for GOLPE or SIMCA

GOLPE and SIMCA are programs which can be used in order to analyse Grid maps statistically.  Further information about GOLPE may be obtained from Tripos Associates or from:

Prof. S. Clementi Universita di Perugia  Via Elce di Sotto 10 06100 Perugia Italy

Telephone: +39-75-5855550 Email sergio@chemiome.chm.unipg.it

and further information about SIMCA may be obtained from:

UMETRI Box 1456 S-901 24 Umea Sweden

Telephone: +46 90 154840 FAX:       +46 90 197685

Users must have been supplied with Programme GRID by Molecular Discovery Limited, and must have valid authorities to use GRID and/or GOLPE and/or SIMCA.  Unauthorized users are liable to prosecution.

>THE GRID:GOLPE INTERFACE

GOLPE allows Users to import GRIDKONT output directly from Programme GRID. Directive LIST would be set equal to 1 during the GRID run if a single Target was being studied.  Alternatively LIST would be -2 when working with a Set of Targets.  In either case the GRIDKONT output from GRID is in binary, and can be read straight into GOLPE. No other values of LIST are acceptable for the GRID:GOLPE combination.

GOLPE can also accept GRINKOUT as input files with the structure of the Target or Targets; i.e.  they can use GRINKOUT files like PDB files.  Therefore if you are studying one Target (LIST=1) you can display the Grid map (GRIDKONT) on GOLPE together with its Target (GRINKOUT).  Of course no statistical computations would be possible with GOLPE, because you would only have one Target in this case.  If you are studying a Set of Targets (LIST=-2) you can still display the structures and maps, and can also do statistics.

>STATISTICAL ANALYSIS WITH GOLPE

When LIST=-2 you may be studying several Targets with several Probes. GOLPE will read and interpret the GRIDKONT file, and will correctly distinguish between Targets and Probes.  It will then ask how you want the data to be presented.  There are two alternative layouts which are discussed below under the Description of SIMCA.  Also see above under the heading:  COLUMN NOMENCLATURE FOR PLS.   You should select the File GOLPE Menu and then:  Import Fields -> GRID-kont.   A new window will appear in which you should select the GRIDKONT file and immediately save it in GOLPE Format with a new filename (e.g. NEWFILE.DAT).GOLPE will ask you to make your choice "by specifying if you want to split the probes so duplicating the objects". If you select "YES" you will have for each molecule a number of objects equal to the number of probes, otherwise you will have an overall number of objects equal to the imported molecules, because each molecule will be described by all the probes values, one probe after the other.

  Opening the DAT file in GOLPE the GRID maps can then be displayed graphically by selecting Plot -> Grid plot -> Present fields.

>THE GRID:SIMCA INTERFACE

The output from GSIM can be read as input by most versions of SIMCA. If you have a copy of SIMCA which apparently does not accept the output from GSIM, we suggest that you:

• Inspect the ASCII output from GSIM, and compare it with the format specified for your copy of SIMCA.

• Then either:

  • edit the WRITE statements in GSIM appropriately, or

  • obtain an appropriate copy of SIMCA from UMETRI, or

  • edit the output file from GSIM.

>POST-PROCESSING GRIDKONT FOR SIMCA

When several Targets have been studied as a Set using several Probes as described above, the GRIDKONT output file must be postprocessed before it can be used as input for SIMCA.  This is done with Programme GSIM which which runs directly from the keyboard.

The executable on Unix systems is called gsim, and this should be typed on a line by itself in response to the Unix Prompt %:

 % gsim

In either case one is asked:

 Type name of your GRIDKONT input file:

and you answer with the name of the GRIDKONT output file generated by GRID for your Set of several Targets with several Probes.  We will use the same example, with FILE.LIST and phenol.pdb phenolate.pdb and pyridine.pdb as described above.

The next questions are then asked:

   How many compounds have you studied:
   3
   How many Probes have you studied:
   4
   Is this an ASCII file which you can read by eye?

and the answer to this question is YES for the present example, but if the binary output option had been used the answer to this question would naturally be NO.  The dialogue then continues:

   Is the 'MESSAGE' the Biological Activity of each compound?   Answer    YES or NO or HELP:

and this question refers to the MESSAGE associated with the compound names in FILE.LIST as described above.  For this first example we suggest that you answer YES, so that the MESSAGE (i.e. the numbers 1.5 -1.5 and 0.5 in FILE.LIST) will be carried forward as a User Defined Column of Y-data to the input file for SIMCA.  (In fact the numbers represent partition coefficients, and not "Biological Activities" in this example).

The exact layout of that input file for SIMCA will be determined by your answer to the next question:

   Do you want to study the compounds?
   Answer YES or NO or HELP

If you answer YES then the input for GOLPE or SIMCA will be arranged like this:

exactly as it was described above under the heading: COLUMN NOMENCLATURE FOR PLS.

An alternative layout is available if you answer NO, and the input will then be like this:

Finally you must type a short file name for the file which is going to be prepared as input for SIMCA.  This name may only contain letters and numbers (No full stop; no file extension; no semi-colon; no generation number) and may only contain 6 characters.   FIRST would be an acceptable file name, and as soon as you type it (and hit the RETURN key) you should get a message from Programme GSIM like this:

  Opening your GRIDKONT file.   Please wait

GSIM has now started to prepare the input for SIMCA, and soon finishes by printing a brief summary description of the Grid.  It should finally display the message:

  FINISHED   Your files are ready for SIMCA

and you have successfully prepared a new file with all the necessary results from GRID in order to start a SIMCA analysis.

>GRID/GOLPE/SIMCA SUMMARY

The GOLPE and SIMCA methods are fully described in their respective literatures which are distributed with the software to authorised Users of those Programmes.  In the present example you have prepared input to GOLPE or SIMCA for three compounds (phenol, phenolate and pyridine).  Each compound was studied using four GRID Probes (C3 OH2 N3+ and O::) which provided the so-called X-values for the statistical analysis.  The output file (GRIDKONT.DAT) from GRID was imported to GOLPE or SIMCA.    The appropriate Y-value (called the MESSAGE in the FILE.LIST) for each compound was also written to the SIMCA file, but you would have had to enter it separately into GOLPE.

PERMISSIONS

EVERY USER OF THE GOLPE OR SIMCA SOFTWARE MUST HAVE OBTAINED DIRECTLY FROM THE SUPPLIER A CURRENT VALID AUTHORITY TO USE IT.  Your supplier should also be consulted for the most up-to-date information.