[PLS-coefficients][PLS-loadings][PLS-partial weights][PCA-loadings][Field sd][Present fields][D-optimal selected var][F. Factorial selected var][Activity-Contributions][PLS-score-Contributions][PCA-score-Contributions][N-level variables][Groups of variables][Objects differences]
All the commands in this submenu start a grid plot application. Follow this link for a full description of its options and commands.
Notice that this is a "tear-off menu", if you click on the dotted line on top of the menu you will transform the submenu in an independent window which will remain open until you specifically close it. This allows a fast access to the plots and is particularly usefull when you plan to open many plots at the same time.
The Grid plot submenu contains commands which plot data in the three-dimensional grid-cage. This plot is only applicable to X-variables previously defined as being of grid type with the File>>>Type of variables>>>Modify command. Notice that grid variables are supposed to be obtained from a grid analysis, and therefore each variable has associated a position in the space around the molecules. There are two ways for representing this kind of data:
Symbol plot
A symbol is plotted for each variable. The size of the symbol is proportional to the absolute value of the data and the color represent the sign (GOLPE uses always light blue for negative and yellow for positive values).
Isocontour plots
GOLPE draws a surface that connects all the points of the grid space having a given value of the data. This value should be provided by the User.
Plot>>>Grid plot >>PLS-coefficients
Plot>>>Grid plot >>PLS-loadings
Plot>>>Grid plot >>PLS-partial weights
Plot>>>Grid plot >>PCA-loadings
First, the User is asked about the component which will be represented in the plot. When there are more that one block of X-grid variables, the User should also choose the field which will be represented in the grid plot.
When the settings are correct, press the OK button. Press the Cancel button to abort the operation.
A grid plot with the values of PLS-coefficients, PLS-loadings, PLS-partial weights or PCA-loadings will be shown.
Plot>>>Grid plot >>Field stand.deviations
This command displays the grid plot of the standard deviations of each variable. It is useful to visualize the regions with maximum X-variance, or with small X-variance around the molecules.
Plot>>>Grid plot >>Present fields
This command displays a grid plot with the values of the X-grid variables (fields) for a given molecule.
It appears a dialog window in which the User can select the molecule the field of which will be plotted and, when the data file contains more than one single X-variables of grid type, the field to plot.
Objects:
A list with all the objects (molecules), identified by their number and their name as they appear in the data file. Click on any item to select this object for the grid plot.
Show field:
The block of X-variables of grid type which will be represented in the grid plot.
When the settings are correct press the OK button. Press the Cancel button to abort the operation.
Notice that in the grid-plot the interaction energies will be represented in different colors: light blue for negative values (attractive interactions) and yellow for the positive ones (repulsive interactions).
Plot>>>Grid plot >>D-optimal selected variables
Plot>>>Grid plot >>FFD selected variables
WARNING: the D-optimal selection list or the F. Factorial selection list will appear only after performing the corresponding variable selection procedure
Displays a grid plot showing the variables selected in a previously made D-optimal preselection or F. Factorial variable selection procedure.
A dialog window like this is presented:
D-optimal selection list
F. Factorial selection list
This list contains all the previous variable selection procedures performed on this data file. Each procedure is identified by a sequential number, the initial number of variables, the final number of variables, the number of components and the hour and date when the procedure was finished. Click on any item to select it. The selected items will be included in the Selection: input field.
Selection:
The selection from the above list is shown in the Selection:input field.
When the OK button is pressed, all the variables selected in the variable selection procedure chosen will be represented in the grid-plot, showing their positions in the three-dimensional grid space around the molecules.
Plot>>>Grid plot >>Activity Contributions
This option allows to visualize, for a given molecule and model dimensionality, how the field values contribute to the activity. The plot is useful to identify the regions of the grid space around a given molecule which contribute most to explain its activity. So, positive areas (yellow) highlight the areas which, in this molecule, contribute to increase its activity. On the contrary, negative areas (light blue) highlight the areas which, in this molecule, contribute to decrease the activity.
It is important to point out that the activity contribution plots are object dependent (molecule dependent), since the plot is obtained as the product of the field value of the selected molecule times the model pseudo-coefficient, for each grid location.
A dialog window like this appears:
Objects:
A list with all the objects, identified by their number and their name as they appear in the data file. Click on any item to select this object for the grid-plot.
dimensionality:
The dimensionality of the PLS model which should be used for the calculation of the activity contributions.
Show field:
The block of X-variables of grid type which will be represented in the grid-plot.
When the settings are correct press the OK button. Press the Cancel button to abort the operation.
Plot>>>Grid plot >>PLS-score-Contributions
Plot>>>Grid plot >>PCA-score-Contributions
These plots highlight the areas of a molecule responsible for its position in the scores plot, with respect to a given model component. This plot might be useful to understand why two objects are different from the point of view of a PCA or PLS model, and for identifying outliers.
It is important to point out that the activity contribution plots are object dependent (molecule dependent), since the plot is obtained as the product of the PCA or PLS loadings multiplied times the values of the field variables, for a given molecule.
A dialog window like this appears:
Objects:
A list with all the objects, identified by their number and their name as they appear in the data file. Click on any item to select this object for the grid-plot.
component:
The component of the PCA or PLS model which should be used for the calculation of the score contributions.
Show field:
The block of X-variables of grid type which will be represented in the grid-plot.
When the settings are correct press the OK button. Press the Cancel button to abort the operation.
Plot>>>Grid plot>>>N-levels variables
This command displays a grid plot showing variables with peculiar behavior and therefore potentially dangerous for the analysis. The potentially dangerous variables should be removed from the data file with the Advanced pretreatment tool. Please see Pretreatment>>>Advanced Pretreatment command for details.
A dialog window like this is shown.
The User can choose to include in the plot one or more of the following types of variables:
inactive variables
Variables with a weight of 0.000. A variable can have a zero weight because of different reasons:
constant variables
Variables which have exactly the same value for every object in the data file and therefore no variance. They usually correspond to areas in the cage far away from the molecular structures or related with the common skeleton used to superimposed all the molecules of the series.
2-levels variables
Variables which take only 2 values in all the data file, one of which appears only in 1, 2 or 3 objects, following one of these three patterns.
|
level 1 |
level 2 |
|
1 object |
rest of objects (n-1) |
|
2 objects |
rest of objects (n-2) |
|
3 objects |
rest of objects (n-3) |
3-levels variables
Variables which take only 3 values in all the data file, one of which appears only in 1 or 2 objects and a second one only in 1 or 2 objects too, following one of these three patterns:
|
level 1 |
level 2 |
level 3 |
|
1 object |
1 object |
rest of objects (n-2) |
|
1 object |
2 objects |
rest of objects (n-3) |
|
2 objects |
2 objects |
rest of objects (n-4) |
4-level variables
Variables which take only 4 values in all the data file, following one of these three patterns:
|
level 1 |
level 2 |
level 3 |
level 4 |
|
1 object |
1 object |
1 object |
rest of objects (n-3) |
|
1 object |
1 object |
2 objects |
rest of objects (n-4) |
|
1 object |
2 objects |
2 objects |
rest of objects (n-5) |
All these 2, 3 and 4 level variables are dangerous because of their skewed distribution: if used to build a PLS model they will take the lead in defining the latent variables, but only because they force the model to explain most of the variance of a few objects with a high leverage, thus leading to spurious and misleading results. We strongly recommend to remove these variables from the data before starting the analysis.
When all the desired kinds of variables are selected press the OK button. Press the Cancel button to abort the operation.
Plot>>>Grid plot >>Groups of variables
This command displays a grid plot which contains all the variables included in one or more of the groups generated with the command X-selection>>>Generate groups or X-selection>>>Region selection>>>Generate regions. If there are more than one block of field type X-variables, two or more windows will appear, one for each field.
The groups are identified by their sequential index and the number of variables contained.
Click in one or more items of the list and then press the Plot button to generate the grid-plot. This dialog window will persist in the screen until the User press the Exit button, thus giving the opportunity to display several grid plots for several groups.
Plot>>>Grid plot >>Object differences
Displays a grid plot representing the arithmetical differences between the values of the field variables of two given objects. These plots are useful for interpret of some models.
A dialog window like this is presented:
The User should select an object from the list Object A and the other from the list Object B and then press the Plot button. Then it will appear one or more grid-plots (as many as X-grid blocks of variables are present) representing the differences between the values of the X-variables for the object A and the values of the X-variables for the object B.
This dialog window will persist in the screen until the User press the Exit button, thus giving the opportunity to display several grid-plots of objects differences.