SAR Table is an app for iOS devices (iPhone, iPod and iPad) which operates on a collection of documents which store SAR data. SAR is a commonly used acronym for structure-activity relationship, which is commonly used in the drug discovery industry to correlate certain chemical functionality with increased or decreased activity against biological targets. The term structure-property relationship is also commonly used, and refers to the same techniques, for chemistry in general. The app is available on the Apple iTunes AppStore.
The arrangement of a series of chemical structures into a table defined by scaffold + values of substituents is a method frequently used to understand the relationship between structure and function, and is especially prevalent in the medicinal chemistry literature:
While arranging data in this form is an effective way to understand the effects of chemical modification, and is equally effective for communicating this information to other scientists, preparing the data in a publication ready form can be a tedious task. Typically this is done by combining a word processor with a chemical drawing package. The table is assembled by drawing the scaffolds and either drawing or abbreviating the substituents (often called R-groups).
This workflow and format separates the presentation and chemical interpretation, and avoiding mistakes with regard to the chemical meaning of the table is entirely the responsibility of the operator. Because there are many opportunities for errors, it is essential to prepare such tables with painstaking care. If the actual chemical data is also going to be used for other purposes, e.g. submitting the molecules to a compound registration system, or performing computational chemistry operations such as QSAR, it becomes necessary to keep a second representation of the data, since these methods all require that each molecule be drawn in full, rather than composed as fragments.
The SAR Table app is designed to address these issues, and to make the process possible on a mobile device.
The SAR Table app defines the eponymous unit of operation as a SAR Table, which describes some number of chemical compounds, which are individually defined by scaffold, substituents and the constructed molecule, which is typically implied. It also holds auxiliary information, such as id, text, and properties (with units). Because this information is stored together in the same datastructure, it is possible to provide a number of features for checking the veracity of the structures, and a provide user interface shortcuts to minimise the amount of operator time required to specify the data. It is also possible to derive presentation-quality output in a number of different styles, which can be used for collaboration and publication.
The app is designed for two scenarios:
While these two scenarios represent opposite circumstances, the workflow requirements are very similar. In both cases the data is stored in a structured manner that is complete, and suitable for creating publication quality output, for retention of records, and for use by all manner of cheminformatics applications.
The next part of this tutorial describes creation of a SAR Table document using a classic series from the dawn of QSAR:
Creating a new table
On the main screen, double-tap on the Create New icon:
The field-setup dialog box will be presented:
There are initially 4 pieces of information showing: Title, Description, Scaffold and Molecule. Scaffold and Molecule are the two default fields that are always present in a SARTable, and these are sufficient for now. The Description is optional, but something must be provided for the Title. Double-tap on it:
Enter a concise description for the document, then press Accept.
Now use the Apply action button:
This will create a new blank SARTable document, and display the icon on the main screen:
To open a SARTable document for editing, double-tap it. Editing the newly created table will show just the field headings, a command bank, and no rows. Use the Add Row action button to create the first row:
Now double-tap on the cell underneath the Scaffold field:
This will open up the structure editor. Use it to draw the underlying scaffold for this data series:
Note the use of 3 non-atom labels: R, R1 and R2. The latter two are used to denote the meta and para substituents that will be used as the table is filled out. Where scaffolds are concerned, any non-element label is considered to be a placeholder for a substituent. These often have labels such as X, E, R, R1, R2, etc.
Accept the structure change. This will, first of all, replace the empty cell with the structure that has just been drawn. But a number of other features come into play. First of all, the app will examine the scaffold and check to see if any of the substituent placeholders happen to be lacking a corresponding field. Since this is a new table, and 3 substituent placeholders have been referenced, none of them have a corresponding field in the table. The following dialog offers to create them automatically:
Press the Create button for the default action, which is to create 3 new substituent fields:
As can be seen, the Scaffold field shows the structure that was just drawn. Each of the substituent placeholders is highlighted in turquoise, which brings attention to the fact that the corresponding substituent field is not defined (on account of being blank). Each of the substituents now has its own field. The values for these fields are blank, and are annotated with a turquoise question mark, which indicates that a definition is expected. Finally, the value of the Molecule field has been created automatically: it has the same structure as the scaffold, but without the substituent placeholders.
The next step is to provide definitions for the substituents. Double-tap the cell for R:
The structure editor appears. Because the value for the substituent was blank, it starts out with a chemical structure that with a single placeholder atom, which is labelled "R", which is the attachment point:
The fragment structure required is phenylcarbonyl:
Accept the structure:
The row now has a definition for the R field. Note that the "R" label is no longer highlighted in the structure of the scaffold.
Note also that the Molecule field has been changed: the phenylcarbonyl structure has been automatically grafted onto the scaffold, so that the auto-created construct is now made up of these two partial fragments.
The remaining substituents describe a para-amino substituent on the scaffold's phenyl ring, which can described by defining R1 as R-H and R2 as R-NH2:
The Molecule field has been updated for each of these changes. Note that the explicit hydrogen atom given for the R1 value has been left out of the construct molecule for clarity.
Now to add the second entry: add a new row, and double-tap the blank scaffold cell:
Instead of activating the structure editor, the double-tap action has taken on a different interpretation. Because the cell is blank, and there are non-blank values available, it has instead invoked the duplicate functionality:
A dialog is presented, which offers a list of each of the unique values from other rows within the current field. In this case there is just one, which is the scaffold being used for the previous row. Select this scaffold and tap Accept:
Apply the same process for R and R1, since they are both the same as for the previous field. For R2, however, the substituent is a methyl group. Rather than double-tapping on the cell, either select it, then tap it again after a pause, or touch-and-hold, which will bring up the context bank. Select the Edit action:
This will open the structure editor, which allows the methyl substituent to be specified:
Adding scalar data
The default and automatically-created fields are usually sufficient for defining the molecular structure fields of a SARTable. Text and data fields must be added using the field editor, which can be accessed from the command bank:
The field editor has action buttons for adding, editing, deleting and moving fields. Activate the Add Property button:
Enter the name of the field:
Press the Accept button, and a new property will be created:
Accept the changes to the field editing session, and the new field will be shown in the table view:
Double-tap the first cell to edit it:
The property editor dialog has 4 sections: Mod, Value, Error and Units. Because this activity value is unitless, the Units part is left blank. The Mod part is a special field for non-numeric notes or modifiers, which are commonly used in tables, such as ">", "<", "n/a", "nd", etc. Keeping this separate from the Value field allows numeric and non-numeric content to be handled with relative grace. The Error section is for entering the standard error of the measurement, if it is available.
The remainder of the table can be finished off using the same steps: add row, copy molecular structures when a diagram already exists or draw a new ones, and enter property data.
Return to the main menu. Select the SARTable that we have finished editing:
Touch-and-hold to bring up the context menu. Select Send Email:
An outgoing email will be composed, which contains the attachment sartable.ds:
The attachment contains the SARTable data, using the XML datasheet format. This is the same format used by other products from Molecular Materials Informatics. These datasheets can be used to interoperate with other mobile apps.
To prepare a SARTable for presentation, touch-and-hold the icon on the main menu, and select Present:
The configuration dialog is displayed:
There are several presentation options available: colour scheme, export format, and the list of fields which should be included in the report.
To view a print-ready version of the report, press the Preview button. A print-ready PDF of the table is prepared and shown:
If a printer is configured, the output can be sent directly to the print device.
From within the preparation dialog, the Email button will initiate an outgoing email, with the presentation-ready file as an attachment:
The default format used when emailing presentation-quality output is PDF, which can be created by the app itself. There are other formats that can be produced with the assistance of a remote procedure call.
HTML + SVG
The table can be rendered as a single HTML file, in which each of the structures is represented by an embedded SVG graphic. SVG (Scalable Vector Graphics) can be rendered by most modern browsers, including mobile Safari:
The resulting HTML document can be used as-is, or modified to suit.
The table can be converted into a Microsoft Word document (.docx), using the Office Open XML (OOXML) variant, which is the default format used from Microsoft Office 2007 onward. The Word document contains an embedded table, and each of the structures is rendered as a vector graphics drawing:
When opened with the appropriate software, the document can be modified to suit. Because the chemical structures are vector graphics, they can be scaled to any size without loss of quality, and are suitable for printing. Content can also be used to compose other documents, e.g. cutting and pasting portions into another Word document, or a PowerPoint presentation.
The table can be converted into a Microsoft Excel spreadsheet (.xlsx):
The resulting spreadsheet preserves the tabular structure of the source material, and renders the structures as vector graphics.
This tutorial has used a straightforward example set to illustrate the basic functionality of the SAR Table app. There are a number of additional features of interest.
When putting together the construct molecule by grafting the substituents onto the scaffold, there are often multiple ways to orient the substituents. Consider the disubstituted benzene scaffold:
The substituent values for R1 and R2 can be grafted onto the scaffold by rotating and translating the fragment as-drawn, or they can be flipped first. Consequently there are 4 possible ways in which the molecule can be constructed. As can be seen, the substituents have been attached in a way that minimises the molecule congestion.
The molecule construction process considers terminal substituents to have 2 possible placement modes, as long as the fragment structure does not contain any stereoactive bonds (i.e. up or down wedges). These degrees of freedom are explored in order to find the least congested layout.
It is valid to include a substituent label more than once when defining the scaffold, e.g.:
In this case, the "X" label is used 3 times, to denote a -CX3 substituent. There is just one X substituent field, and it is used as many times as necessary when generating the construct molecule. Note the label in the bottom right corner, "x3".
Multidentate substituents are handled. Consider the following case:
The scaffold has two substituents, R1 and R2. The second row uses one of the substituent fields - R1 - to draw a substituent which has both of these attachment linkages. The R2 field shows a tilde, to indicate that there is no need to provide a substituent value, since the value of R1 covers it by overlap.
Note that the construct molecule is generated by fusing and aligning these two fragments. It is important to draw multidentate substituents using a style that can be readily joined by rigid rotation.
For multidentate substituents, it is important to make sure that the attachment labels match. In this example, the R1:R1 and R2:R2 labels link up nicely. If the labels do not match, the construct molecule will not be created properly. This is in contrast to single-attachment substituents, which can use an arbitrary label for the attachment placeholder.
The construct molecule, which usually has the field name Molecule, is normally automatically generated. This can be overridden, however: if the field is edited explicitly, the manually-provided structure will be used, and the status of the structure will be set to locked. This means that modifying the scaffold or substituent fields will not cause the construct molecule to be regenerated. The operator must take care to ensure that this does not result in an inconsistency. The cell can be unlocked using the context menu, which will return to the default state of being automatically redrawn whenever the constituent fragments are modified.
This feature comes in useful when the automatically generated structure diagram is not suitable, or when adding structures to a table which are not part of the series, i.e. not based on a common scaffold.
Inline abbreviations are preserved during the molecule reconstruction process, so it is appropriate to use them for substituent definitions:
This is visually more apparent when one of the construct molecules is viewed in the editor:
The diagram on the right shows the display when the tBu substituent is the current atom, exposing the structure of the inline abbreviation.
For cases when the construct-molecule is defined, and the basic core of the scaffold is known, the matching can be done automatically. For a detailed discussion, see: SAR Table: Scaffold Matching.
Property fields can be associated with a scheme, which provides information about the units, the useful range, and a colour-interpolation scheme which can be used to colour-code property values. For a detailed discussion, see: SAR Table: Schemes.
The SAR Table app provides a datastructure for storing structures and property information, in a format that maintains the assigned fragments together with the whole molecule, and associated property and text data. The user interface provides a number of convenience shortcuts, and automates as much of the process of creating these tables as possible. The app is designed for creating new tables prior to communicating structure-activity data, or for recreating the underlying data from literature publications.