| MetaSite user manual | ||
|---|---|---|
| <<< Previous | Next >>> | |
Chapter 4. Using the MetaSite GUI
MetaSite is launched from a shortcut in the Windows start menu.
Linux users can start MetaSite by simply typing its name in any terminal window, followed by no argument:
[ric@proton ~]$ metasite & |
Depending on the current phase in a work session, on the availability of imported substrates, and on the availability of the corresponding prediction results, only the appropriate pages are activated. Pages unaffected by the process performed are disabled until required.
Since no data can be made available at program startup, the user is presented with the MetaSite3 welcome page. This page briefly presents some information on "getting started", shortcuts to the user manual and tutorials, and a link to the Molecular Discovery Ltd web site.
4.1. Importing substrates
Substrate structures may be either imported into the document database in several different electronic formats (SDF, MOL, MOL2, SMILES...), directly sketched in the MetaSite GUI, or otherwise pasted from ISIS/Draw.
| Warning |
Pasting from ISIS/Draw is only available in the Windows version. |
To import, either click on the File->Import menu or the corresponding Import toolbar buttons:
As part of the import process, structures are assigned the object name they will be associated with in the internal database and in the graphical user interface. Object naming may be based on different policies, some of which only apply to the appropriate input formats:
UID: this is the most basic naming policy and is the default setting. Objects are assigned a name that looks like Object N where N is a unique, increasing, and numerical identifier in the internal database.
Filename + ordinal number: objects are assigned a name that is based on the name of the corresponding input file. Should the file contain multiple structures, the resulting object name is appended with a numerical suffix indicating the structure position in the file.
by Structure name: the object is assigned the same name as the corresponding molecular structure in the input file, should there be one. Should no structure name be available from the input file, the policy uses UID instead.
by tag (SDF only): the contents associated with the SDF tag selected are used to name the object. Should the selected tag for the object being imported not be available, the policy uses UID instead.
Name as: when import of one single object is implicit from the selected input format or source (for example, when the user enters a single SMILES string, or a structure has been sketched), the user may directly enter a suitable name.
Figure 4-3. The objects import dialog shows a selected naming policy. The user is importing substrate structures from an SD file. The name of the objects in the document dataset will be given by the contents of the specified SDF tag (Molname).
Import of SMILES list files is supported for three specific list formats:
Figure 4-4. Objects import dialog for SMILES list. The names of the objects in the created dataset are optionally provided by additional fields in the list format.
<SMILES string> <structure name>: each line in the list file is expected to contain two blank-separated fields corresponding to the SMILES string and object name, in this specific order.
<SMILES string>: each line in the list file is expected to contain one single field, corresponding to the SMILES string. Objects are assigned a name based on the UID policy.
<structure name> <SMILES string>: each line in the list file is expected to contain two blank-separated fields corresponding to the object name and SMILES string, in this specific order.
Editing dataset objects as new: Double-clicking any depiction of a molecular structure appearing in the MetaSite GUI opens the 2D molecular editor that shows the same structure, which is also ready to be imported as a new dataset object. This may be very useful when manually importing a family of structurally related compounds or when the user wants to explore how small changes to a structure may affect the predicted results.
4.1.1. The Dataset View
Users may browse the imported dataset from the Dataset tab page at any time. This view displays the molecules under study in tabular form, each new row corresponding to a newly imported molecule, and is always active when the document is not empty.
The first column in the table (Object) reports the 2D graphic depiction of the compounds. Then the remaining columns represent additional attribute data associated with the dataset objects, and are only reported when the input file provides the related information. If the input file is an SD file, each additional column contains the information of the corresponding SDF tag in the input file. If any molecules lack tag information, the corresponding cells are blank.
The user may manipulate the existing dataset by means of a context menu access by right-clicking on any point of the Dataset view area:
A sliding bar is placed at the bottom of the table and can be used to split the table horizontally, which also makes it possible to visually compare structures very far apart from each other in a large table.
4.1.2. Import directives
Molecular structures are preprocessed during import: some data required for later computations are produced in advance and the user may choose to alter the state of substrate protonation. MetaSite is capable of generating three-dimensional structures when needed, and of sampling the associated conformational space. Consequently, an input 3D representation is not generally necessary.
Once the input data is selected and the naming policy is set, users may want to access the Directives page of the Objects import dialog box, and optionally configure the import process:
The default settings of the Directives page are configured according to the application preferences and settings, but users may occasionally wish to modify its contents in order to override these settings for a specific import operation.
Protonation policy may be set as follows: as is (protonation in the input structures remains unchanged), neutralize (a suitable neutral form will be imported whenever possible), and normalize at pH (the most abundant species at the given pH value will be imported in place of that stipulated as input). In order to perform these tasks, MetaSite internally produces the required pKa estimation, so once the protonation policy has been set, no additional user intervention is required.
| Warning |
We recommend using the neutralize option to import chemical structures. In fact, Ri and Mi components have been optimized for neutral structures. |
4.2. Site of Metabolism prediction
On selecting the Tools menu, then Site of metabolism prediction or clicking on the corresponding toolbar button, a dialog box appears on the screen prompting the user to select the dataset objects and cytochrome models to be used for the site of metabolism prediction procedure:
The user has the option of either selecting one of the consensus models (LIVER, SKIN or BRAIN) or one or more of the specific CYP isoform models that are available in the previously installed enzyme libraries.
Once the hot-spots prediction data is made available, the corresponding tab pages in the graphical interface are enabled.
4.2.1. The SoM Classic View
The SoM Classic View provides the most detailed presentation of the Site of Metabolism (SoM) prediction results for a selected substrate/CYP pair.
This view is similar in concept and overall organization to the graphical user interface in earlier versions of MetaSite, but now a new implementation makes it more user-friendly and easier to access the most relevant information.
4.2.1.1. Analysis of the SoM prediction results
The tab page may be divided into three areas. Two auxiliary areas are represented by the list boxes on the left (substrate and CYP selection) and the text window in the upper right area (displaying numerical data corresponding to site selections performed by the user), whereas the user's attention is usually mainly focused on the wider central area:
This latter area is mainly composed of two interactive graphical panels with controls surrounding it. Two pairs of buttons are located in the upper corners of this central area so that the user can toggle between the two panels to view either the substrate depiction or the bar graph:
In the default configuration of the GUI, the structure of the substrate is shown in the right-hand panel and a bar graph is displayed in the panel on the left. Atoms in the graphic depiction and bars in the graph are in a 1-to-1 relationship and their selection is synchronized. The information reported in both panels corresponds to the site of metabolism prediction (as indicated in the pulldown above each panel). The sites in the 2D graphic depictions and the bars in the graph are colored according to a stepwise color gradient and assigned a shade that depends on the value of the prediction for each site. The predicted hot-spots are therefore highlighted in the graphic depiction and correspond to the highest bars in the graph next to it.
A vertical, colored bar is placed between the two panels. The buttons at either end of this bar are used to customize the base colors for the gradient used in the GUI. These color settings may be made permanent from the Preferences and settings dialog box.
4.2.1.2. Comparing the hot-spots prediction with the substrate exposure
Two different main contributions are taken into account in the computation of a site of metabolism prediction: probability of exposure towards the heme group, and chemical reactivity. Although users are often interested in the most probable sites of reaction, these are not always the only sites in the substrate structure characterized by related exposure to the heme.
There are two identical frames below each one of the two graphic panels introduced by the previous section. The reactivity correction components in the results displayed can be configured in these two frames.
Settings in the two Reactivity Correction frames are synchronized by default so that a change in one of them applies equally to both. This saves the user some unnecessary GUI manipulation in most of the usual conditions where the two graphical panels provide different representations for the same data.
Clicking on a button located between the two frames deactivates this synchronization so that the two panels can be set independently. This is often useful when both graphic panels show the substrate structure and hot-spot distributions, referring to different levels of reactivity correction, so that they may be visually compared.
Figure 4-6. Exposure analysis The estimated exposure to the heme (on the left) is compared with the full SoM prediction (on the right). The methylaminic group (CH3NHR), predicted as the most likely site of metabolism (right), doesn't correspond to the portion of the substrate structure with higher estimated exposure to the heme (left). The ArCl group is highly exposed to the heme, but not so reactive.
4.2.1.3. Structural contributions to heme exposure
There is a combo-box above each graphic panel that reports the data currently being displayed. Both boxes read Site of metabolism by default, but an additional set of data is available to the user, namely Structure contribution.
Structure contributions provide information about which part in the substrate structure are the most relevant in determining the estimated exposure for a selected site and are most often graphically represented in association with the substrate depiction. In the GUI, the atoms making the greatest contribution are highlighted by colored spots, and the site affected in the substrate is marked by the SOM label.
4.2.2. The SoM Multiple View
The SoM Multiple view provides a tabular representation of the available site of metabolism prediction and allows visual comparison of the results produced for a series of compounds and a set of different CYP models
A frame in the lower portion of the page supports global setting of the reactivity correction level. The Export... button allows exporting the reported data either as a HTML or an SD file. File format selection is performed when a suitable file name is provided from inside the dialog issued on button activation.
A sliding horizontal bar, by default placed at the bottom of the table, may be used to split the table in two portions, so that rows far away from each other in the whole table are easily compared.
4.3. Metabolites identification
To start a Metabolite ID prediction, either select Metabolites identification from the Tools menu or click on the corresponding toolbar button to open an initial dialog box in which the user selects the dataset objects and cytochrome models input.
For each substrate selected, later processing will produce a set of chemically accessible metabolites, ranked according to the results from a hot-spot prediction based on the CYP models selected.
Dependency on SoM prediction: Metabolite identification provides a list of metabolites ranked according to a corresponding prediction of the substrate hot-spots. The user is not expected to execute this prediction prior to executing metabolite identification. If not already available, MetaSite will automatically run the required site of metabolism prediction.
The parameters controlling the metabolites generation step can be customized in a second dialog window.
An initial group of settings controls metabolite filtering options:
Minimum mass: metabolites with a mass below the assigned value are discarded. A null threshold value disables this filtering mechanism.
Ignore metabolites stereochemistry: although MetaSite may keep track of the stereochemistry in the metabolites generated, this is not always useful if the results from predictions are to be compared with experimental data obtained, for example, from mass spectra where this factor is not relevant and stereoisomers would appear as replicated entries in a ranked list. When this option is set, metabolites differing only in their tetrahedral stereochemistry are treated as the same compound.
Ignore redundant metabolites: when a reaction implies bond cleavage (e.g. dealkylation) more than one single metabolite is often produced from the same reaction pathway and the smaller ones are not associated with any additional information. When this option is set, metabolites with a mass percentage (relative to the parent substrate) below the given threshold are considered to be redundant and discarded.
Metabolites generation options: These settings may be made permanent in the Preferences and settings dialog box, which is accessed from the Edit menu.
4.3.1. Metabolites View
The Metabolites view provides a detailed report of the metabolites prediction results for a substrate/CYP pair selected.
The leftmost part of the page contains two list boxes from which the substrate and CYP model are selected. The remaining portion contains a graphic interactive panel with two tables in its upper part and a bar graph at the bottom.
The graphical panel may either display a bar graph (the default mode) or a graphic depiction of the substrate structure. The bar graph represents metabolites rankings; the height of each bar in the graph is proportional to the score of the corresponding metabolite.
Metabolites sorted according to the computed score are reported in a table on the left, together with the monoisotopic mass (MIM) and predicted LogP and LogD values (the latter computed at pH 4, 7 and 9). Reaction mechanisms responsible for the production of each metabolite from the parent substrate are also reported. Above the metabolites table is a smaller single-row table that reports similar data for the substrate compound.
Finally, the bar graph at the bottom of the page provides estimated relative retention times for the metabolites produced.
4.3.2. Metabolite ID Summary View
The Metabolite ID Summary view provides a tabular representation of the available metabolite predictions from which comparison can be made of the most likely metabolites produced for a series of compounds and a selected CYP model.
The first column in the table reports the substrate name and structure. The structures of the most likely metabolites, together with the corresponding MIM values, are shown in the additional columns. There is a pulldown in the upper left-hand corner of the page from which the user can select the enzyme model determining the metabolites rankings. The user may set the maximum number of metabolites shown for each substrate in a spin-box (labeled Max # of metabolites) placed above the table.
In the bottom left-hand corner of the page is a frame where the level of reactivity correction in the underlying SoM prediction can be selected. Finally, the Export... button in the bottom right-hand corner exports the reported data as either an HTML or a CSV file.
A sliding horizontal bar placed by default at the bottom of the table may be used to split the table in two parts so that rows in the table far apart from each other can be compared easily.
4.4. Mechanism Based Inhibition
| Warning |
This feature is activated by a specific license and is not therefore available in a base configuration. |
When the Tools menu is selected and then Mechanism based inhibition (or the corresponding toolbar button is clicked) a dialog box opens prompting the user to select the dataset objects and cytochrome model that will be input in the MBI prediction procedure:
Please note that in contrast to the site of metabolism prediction, identification of mechanism based inhibitors is currently only available for the CYP3A4 isoform and a reduced choice of enzyme descriptors is therefore available from this dialog box.
Selected objects are then sequentially processed and once this has made MBI prediction data available, the corresponding tab page is enabled.
4.4.1. The MBI View
The MBI View provides a tabular report of the available mechanism based inhibition data, and this can be used for visual comparison of the prediction results obtained for the compound/CYP pairs selected.
Each row in the table collects the results for a given compound. The compound name is reported in the first column and MBI prediction results for each CYP model are reported in graphic form in additional cells. The 2D structure of the compound is shown in each one of these cells by a colored flag that indicates whether or not the compound is classified as mechanism based inhibitor (Red), non mechanism based inhibitor (Green), or uncertain (Yellow). Moreover, colored spots on the atoms highlight the fragments of the compound structure that are more likely to cause the reactive metabolite that will lead to mechanism-based inhibition.
The user can use a frame labeled Mbi Filter in the lower left-hand corner of the page to enable or disable the view of compounds based on the estimate of their MBI character.
The Export... button in the lower right-hand corner exports the data reported to external files. The file formats that are currently supported are HTML, SDF and CSV.
A sliding horizontal bar, by default at the bottom of the table, can be used to split the table in two parts so that rows far apart from each other in the table can be compared easily.
4.5. Importing cytochromes
| Warning |
This feature is activated by a specific license and is not therefore available in a base configuration. |
MetaSite is provided with an enzymes library already containing the descriptors for the most relevant human cytochromes. Descriptors for additional library entries, either produced from proprietary protein models or for isoforms that are not already available in the distributed library may be automatically created starting from an input PDB file.
Once created, custom CYP descriptors are most often stored into a separate, user's library file. If the path to this additional library file is defined the software will access both this file and the system one transparently, as if they were a single collection of enzyme descriptors.
To define a new library file it is sufficient for the user to enter a suitable file path for it in the corresponding entry in the Paths page of the Preferences and settings dialog. If the file doesn't yet exist, MetaSite will create it on first use.
The creation of a new enzyme library item is started by selecting the CYP import entry from the Tools and managed by the following dialog window:
User is expected to fill the following fields:
Input PDB file path: the path to the file containing the protein structure.
CYP Label: The name that the software will use in referring the produced descriptors. Please note that the names of the items in a collection should be unique.
User's Enzyme library path: this path is by default retrieved from the application settings.
| <<< Previous | Home | Next >>> |
| Installation | Command-line interface |