Ligandscout Tutorial Cards 493i1o
This document was ed by and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this report form. Report 3l3c15
Overview 3z723u
& View Ligandscout Tutorial Cards as PDF for free.
More details 2i4a6q
- Words: 2,882
- Pages: 8
LIGANDSCOUT TUTORIAL CARD 1
Creating a pharmacophore from a single protein-ligand complex Experience level: basic Time needed: 5 minutes
Views
Sequence
oo Macromolecule view oo Active site view
oo PDB file using oo oo oo oo
Controls
oo Ligand box 4-letter code oo Create pharmacophore Discover the correct (button or menu) ligand oo OR: Create pharmacoFocus on active site and phore for Catalyst/MOE/ switch to active site view Phase Check and eventually correct ligand chemistry Create a structure-based pharmacophore
Advanced controls (opt.) oo oo oo oo oo oo
Change bond type Change atom type Create new bond Move to core Move to environment Compound switching
Description: Type ‘1ke6’ in the upper right area of the screen and press the button ‘ 1ke6’ [1]. The protein will be ed and displayed (this is the ‘macromolecule view’) [2]. Click on the yellow box within the protein representing the ligand - an animated zoom begins ending in [3]. Since ligand structures in the PDB only contain incomplete information, you should always check whether all bonds are typed correctly. Bond types can be changed by selecting a bond (by clicking on it either in the 2D or 3D view) and by using the retype bond button [5] or the keys 1, 2, or 3. Once the ligand is chemically correct, create a pharmacophore by pressing Ctrl-F9 (Cmd-F9 on OS X) [4]. Where to go from here: oo Pharmacophore modeling: Creating shared and merged feature pharmacophores oo Customizing pharmacophore creation preferences oo Virtual screening
1
2
3
4
5
LIGANDSCOUT TUTORIAL CARD 2
Pharmacophore modeling: Create shared & merged feature pharmacophores using chemical features from three PDB structures Experience level: medium Time needed: 15 minutes
Views
Sequence
Controls
Advanced controls (opt.)
oo Macromolecule view oo Active site view oo Alignment view
oo oo oo oo
PDB File ‘1ke6’ Check & correct ligand Create pharmacophore Add ligand and pharmacophore to alignment view oo Repeat the last 4 steps for ‘1ke7’ and ‘1ke8’ oo Create a ‘shared feature pharmacophore’ from the 3 pharmacophores oo Align the three ligands to the new pharmacophore
oo Ligand box oo Create pharmacophore
oo Set reference element oo Center all structures oo Alignment context menu
(button or menu)
oo Add molecule to align-
ment
oo Add pharmacophore to
alignment
oo Generate shared feature
pharmacophore (by features) oo Generate merged feature pharmacophore
Description: Type ‘1ke6’ in the upper right area of the screen and press the button ‘ ‘1ke6’. The protein will be ed and displayed in macromolecule view. Click on the yellow box to focus on ligand ‘LS2201’. Check the ligand for its correctness. Press ctrl-f9 to create a pharmacophore (cmd-f9 on OS X). Use the data exchange widget and select ‘add molecule to alignment view’ [1a] and the molecule will be added to the alignment view, increasing a counter in the corresponding tab. Use the data exchange widget and select ‘add pharmacophore to alignment view’ [1b] to do the same with the pharmacophore. Repeat these steps with ‘1ke7’, and ‘1ke8’. Finally, you should have six elements in the alignment view. Click on the ‘Alignment’ tab to work with them. Select the three pharmacophores in the alignment list using Ctrl-click or multiple selection [2], and press the ‘Generate shared feature pharmacophore (by features)’ button [3]. A new pharmacophore called ‘Shared [LS2201-1, LS3201-1, LS42-1]’ is appended to the list. Mark this element and make it the reference element using the ‘Set reference’ button, which will change the display style of this element to red and bold. Now use the ctrl key (cmd on OS X) to additionally select the molecules identified by the blue icons [4]. Additionally, hide the three pharmacophores derived from 1ke6, 1ke7, and 1ke8 by clicking on the eye symbols on the right next to these alignment entries [4]. Now press the ‘align’ button. You will see the three ligands aligned to the shared feature pharmacophore derived from the three PDB complexes [5]. Where to go from here: oo Customizing pharmacophore creation preferences oo Virtual screening in LigandScout
1 2
3
4
5
LIGANDSCOUT TUTORIAL CARD 3
Virtual screening and comparing active/inactive hit retrieval rates Experience level: advanced Time needed: 15 minutes
Views
Sequence
Controls
Advanced controls (opt.)
oo Macromolecule view oo Active site view
oo Create pharmacophore
oo Ligand box oo Create pharmacophore
oo Virtual screening with
for virtual screening oo Initiate virtual screening oo Refine pharmacophore model oo Repeat virtual screening
omitted features
(button or menu) oo Initiate virtual screening
Description: Virtual screening aims at the maximum enrichment of active compounds in a hit list. Therefore, such methods are usually validated by assessing the accuracy of discrimination between actives and decoys. A good pharmacophore model will be able to identify a significant portion of known active compounds and as few decoys as possible. In this tutorial, two different types of precalculated multiconformational databases are available: A library containing a selection of known active compounds and a library of decoys. Type ‘1ke7’ in the upper right area of the screen, press the button ‘ 1ke7’ . Create a pharmacophore of the proteinligand complex as described in the LigandScout tutorial sheet 1 ‘Creating a pharmacophore from a single protein-ligand complex‘ [1]. Perform virtual screening using a multi-conformational database with known active compounds (filename ‘cdk2ligands.ldb’) for screening using the ‘Screen pharmacophore against external library’ item located in the ‘Pharmacophore’ menu [2]. Virtual screening will be initiated after selecting the library file (press the butten with the three dots to open a file dialog) and the number of omittable features [3]. Since the initial pharmacophore was derived from all the interactions of only one single ligand, it is rather restrictive and retrieves only few hits. Downsize the number of features by deleting less important features [4] or increase the tolerances of the features by selecting a feature and then choosing ‘Pharmacophore’->’Increase Selected Feature Tolerance’ from the menu. Compare the ratio of active compounds and decoys (library filename ‘cdk2-decoys.ldb’) retrieved by your pharmacophore in order to estimate its predictive power [5]. Further adapt the pharmacophore in order to improve the ratio of actives and decoys obtained by virtual screening: Remove non-essential chemical features and exclusion volume spheres in order to increase sensitivity; adapt the size of the feature spheres by pressing ‘ctrl-shift-plus’ or the corresponding menu entry (‘Pharmacophore - Decrease Selected Feature Tolerance’) or use the pharmacophore from tutorial card 2 using the data exchange widgets [6]. Where to go from here: oo Pharmacophore modeling: Creating shared and merged feature pharmacophores oo Customizing pharmacophore creation preferences oo Analyzing screening results
1
4
2
3
5
6
LIGANDSCOUT TUTORIAL CARD 4
Analyze LigandScout pharmacophore screening results Experience level: medium Time needed: 10 minutes Prerequisites: Virtual screening hit list
Views
Sequence
Controls
Advanced controls (opt.)
oo Library view (spread-
oo Create or import
oo Library view (Spread-
oo Calculation of Gaussion
sheet) oo Hierarchy view
screening results oo Sort virtual screening hit list oo Inspect and compare hits oo Discover controls of the library view
sheet) oo Viewer controls
Shape Similarity Score
Description: Perform virtual screening to obtain a hit list as described on tutorial card 4 ‘Virtual screening and comparing active/inactive hit retrieval rates’ or insert external screening results from a library file via the ‘File - Insert’ menu entry. Results of virtual screening are depicted in the library view [1]. Sort the hit list by a single click on the ‘Pharm. Fit Score’ table header [2]. You may invert the sorting order by an additional click. Select a compound of interest by single click on the corresponding table row to be depicted in the active site view together with the original ligand [3]. Multiple compounds are selected when holding shift or ctrl (cmd on MAC OS X) while clicking. Keep compounds visible independent from the current selection state by clicking the eye symbol in the table left to the molecule name in the library view spreadsheet [4]. The visibility of the core molecule (i.e., the original ligand) and other items in the active site view is toggled with the eye symbol located in the hierarchy view on the right side next to the molecule name [5]. Select a custom color for the core molecule using the square icon next to the eye icon in the hierarchy view. Calculation of the Gaussian Shape Similarity Score offers further functionality for scoring. Use the ‘Calculate Gaussion Shape Similarity Score’ function located in the ‘Library’ menu for calculation. A new property ‘Shape Tanimoto Score’ will be added in the library view and you can use the spreadsheet functionality for sorting and ranking as described above. Where to go from here: oo Pharmacophore modeling: Creating shared and merged feature pharmacophores oo Customizing pharmacophore creation preferences
2
3
4
5
1
LIGANDSCOUT TUTORIAL CARD 5
View and rescore docking poses with pharmacophores Experience level: medium Time needed: 10 minutes Prerequisites: docking poses from external sources Views
Sequence
Controls
oo Library view (spread-
oo oo oo oo oo oo
PDB File Switch to active site view Import docking poses Analyze docking poses Rescore docking poses Align docking poses to pharmacophore oo Discover the viewer controls of the library view
oo Library view (Spread-
sheet) oo Hierarchy view
Advanced controls (opt.)
sheet)
oo Viewer controls
Description: ‘1ke7’ and create the pharmacophore of the protein-ligand complex. Import the corresponding docking poses stored in MDL SDF file format (tutorial filename ‘1ke7-docking.sdf’) via the ’File’ - ‘Insert’ menu entry [1]. Make sure that the ‘Align Molecules Automatically to Shown Pharmacophore’ option located in the ‘Library’ menu is deactivated [2], which ensures that the coordinates of the docking poses relative to the protein are preserved. The compounds are imported to LigandScout including all properties given in the SD file, such as the docking score. Sort the list of docking poses by a single click on the table header corresponding to the docking score [3]. Invert the sorting order by an additional click if necessary. LigandScout offers four different options for rescoring docking poses based on the 3D pharmacophore that is currently visible in LigandScout: The calculation of the pharmacophore fit of imported docking poses considering or ignoring excluded volume spheres, and the calculation of the pharmacophore fit of the docking poses after alignment to the pharmacophore considering or ignoring excluded volume spheres. Alignment to the pharmacophore before scoring may be useful to optimize docking poses [4]. All four modes are accessed via the ‘Library’ menu. You can modify the pharmacophore by deleting features, adding excluded volumes, etc. to reflect the important features of the binding site that should be fulfilled by the highest ranked hits. Alignment of imported docking poses to the shown pharmacophore is performed automatically upon selection of a spreadsheet line (i.e. a single docking pose) if the ‘Align Molecules Automatically to Shown Pharmacophore’ option located in the ‘Library’ menu is activated. Poses may also be aligned to the pharmacophore using LigandScout’s alignment algorithm using the ‘Align’ button located in the upper right corner of the library view [5]. Where to go from here: oo Customizing pharmacophore creation preferences
1
2
4
3
5
5
LIGANDSCOUT TUTORIAL CARD 6
Espresso: Create ligand-based 3D pharmacophores Experience level: intermediate Time needed: 15 minutes
Views
Sequence
Controls
Advanced controls (opt.)
oo Ligand-based Modeling
oo Import ligands oo Cluster ligand-set in 3D oo Generate ligand-based
oo 3D Clustering oo Generate ligand-based
oo Use different settings for
pharmacophore model
pharmacophore model (button or menu)
conformational model generation oo Use different settings for clustering oo Use different settings for ligand-based pharmacophore model generation
Description: Open the smiles file containing your ligands (‘cdk2.smi’) by choosing the ‘File-Open’ menu. The ligands are imported and shown in 2D on the right side of the screen [1]. Set all candidates for the model to ‘training set’ by choosing ‘Ligand-Set -> Flag Selected Molecules as Training Set’. Cluster the ligand set according to their 3D pharmacophore characteristics using the ‘cluster’ button on the bottom of the 3D view [2]. Keep the default setting in the following dialog and start the clustering process by pressing the ‘OK’ button. Espresso will create a new column in the ligand table called ‘Cluster ID’. Sort the table using this column by clicking on the column header ‘Cluster ID’. Set all molecules with cluster ID ‘1’ to test set by using the table controls [3]. Now click on the button ‘Create ligand-based pharmacophore”’[4] using the default values to start pharmacophore generation. A ligand-based model will be created and displayed in the 3D view. Feature patterns [5] in the table indicate, which features are met by which ligand. A click on a colored square indicates, which feature is linked to the corresponding square. The list on the right lower corner of the screen shows several pharmacophore solutions: by selecting a different solution, it is shown in the 3D view with the ligands aligned accordingly [6]. You can use the data exchange controls on the upper right corner of the 3D view to move the currently shown pharmacophore to other views (e.g. screening) or save it using the ‘File -> Save’ menu and selecting a file type that represents pharmacophores. Where to go from here: oo Align structure-based pharmacophore models with ligand-based ones oo Virtual screening in LigandScout
1
2
3
4
5
6
LIGANDSCOUT TUTORIAL CARD 7
Screen several databases using the “Screening Perspective” Experience level: intermediate Time needed: 5 minutes
Views
Sequence
Controls
Advanced controls (opt.)
oo Screening view
oo Add pharmacophore
oo Load Screening Database
oo none
model to screening view oo Load virtual database oo Perform virtual screening
(Button) oo Perform Screening (Button)
Description: In contrast to the virtual screening procedure in the ’Structure-based Modeling’ perspective which is designed for pharmacophore model refinement with respect to ligand-binding pocket properties, the ‘Screening’ perspective provides a fast and convenient way of performing multiple virtual screens with different screening databases against your final pharmacophore model, which can be useful for validation. Import your favorite pharmacophore model (e.g. the one created from tutorial card 1 or 6) into the ‘Screening’ perspective either by using the ‘Exchange Data Widget’ [1] or the ‘Open’ entry in the ‘File’ menu. Click on the ‘Load Screening Database’ button [2] to load the screening database ‘cdk2-ligands.ldb’. Enable the loaded screening database for screening by clicking on the grey button left to the screening database entry [3]. Now, select the pharmacophore model from the ‘Pharmacophores’ list to make sure it is shown in the 3D view. Next to the ‘Load Screening Database’ button, the ‘Perform Screening’ and ‘Delete Pharmacophore’ button will be enabled. Click on the ‘Perform Screening’ button to initiate the virtual screening. View the results in the table at the bottom of the screen [4]. You can add more databases and pharmacophores for further virtual screening experiments. Please make sure that only one pharmacophore is visible before you start screening. Where to go from here: oo Analyze screening results oo Advanced screening: ROC curves and model combination
1
2
4
3
LIGANDSCOUT TUTORIAL CARD 8
Advanced screening: ROC curves and model combination Experience level: advanced Time needed: 15 minutes
Views
Sequence
Controls
Advanced controls (opt.)
oo Screening view
oo Add pharmacophore
oo Load Screening Database
oo Boolean Expression
oo oo oo oo
models to screening view Load virtual database Perform virtual screening Plot ROC curve Combine models by ‘OR’
(Button) oo Perform Screening (Button) oo Plot ROC Curve (Button)
(Field)
Description: In contrast to the virtual screening procedure in the ’Structure-based Modeling’ perspective which is designed for pharmacophore model refinement with respect to ligand-binding pocket properties, the ‘Screening’ perspective provides means for statistical evaluation of screening performance. Import the structure-based pharmacophore models from PDB entry 1ke7 (tutorial card 1) and the ligand-based CDK2 pharmacophore (tutorial card 2) into the ‘Screening’ perspective either by using the ‘Exchange Data Widget’ [1] or, alternatively, load a saved version using the ‘Open’ entry in the ‘File’ menu. Click on the ‘Load Screening Database’ button [2] to load the screening database ‘cdk2-ligands.ldb’, and click the same button again and load the database ‘cdk2-decoys.ldb’. Click once on the tri-state selector left to the database names to mark ‘cdk2-ligands.ldb’ as a database containaing actives (green database symbol), and twice on the tri-state slector left of ‘cdk2-decoys.ldb’ to mark this database as decoys (red database symbol) [3]. Now, select the pharmacophore model ‘1ke7’ from the ‘Pharmacophores’ list to make sure it is shown in the 3D view. Click on the ‘Perform Screening’ button to initiate the virtual screening [4]. After screening has finished, press the ‘Plot ROC Curve’ button [5] to plot a ROC curve [6]. Now select both pharmacophores in the ‘Pharmacophores’ table. Numbers in curly brackets (“{1} and {2}”) indicate the number of the pharmacophore. Enter “1 or 2” into the boolean expression field [7] indicating that a virtual hit in this screening run should either fit into the first (1) or to the second (2) pharmacophore. Press the screening button [4] again to re-screen, and the ROC curve button [5] again to compare the virtual screening statistics. View the virtual hits in the table at the bottom of the screen [8]. You can add more databases and pharmacophores for further virtual screening experiments. Where to go from here: oo Analyze screening results
1
2
3
5
6
7
8
4
Creating a pharmacophore from a single protein-ligand complex Experience level: basic Time needed: 5 minutes
Views
Sequence
oo Macromolecule view oo Active site view
oo PDB file using oo oo oo oo
Controls
oo Ligand box 4-letter code oo Create pharmacophore Discover the correct (button or menu) ligand oo OR: Create pharmacoFocus on active site and phore for Catalyst/MOE/ switch to active site view Phase Check and eventually correct ligand chemistry Create a structure-based pharmacophore
Advanced controls (opt.) oo oo oo oo oo oo
Change bond type Change atom type Create new bond Move to core Move to environment Compound switching
Description: Type ‘1ke6’ in the upper right area of the screen and press the button ‘ 1ke6’ [1]. The protein will be ed and displayed (this is the ‘macromolecule view’) [2]. Click on the yellow box within the protein representing the ligand - an animated zoom begins ending in [3]. Since ligand structures in the PDB only contain incomplete information, you should always check whether all bonds are typed correctly. Bond types can be changed by selecting a bond (by clicking on it either in the 2D or 3D view) and by using the retype bond button [5] or the keys 1, 2, or 3. Once the ligand is chemically correct, create a pharmacophore by pressing Ctrl-F9 (Cmd-F9 on OS X) [4]. Where to go from here: oo Pharmacophore modeling: Creating shared and merged feature pharmacophores oo Customizing pharmacophore creation preferences oo Virtual screening
1
2
3
4
5
LIGANDSCOUT TUTORIAL CARD 2
Pharmacophore modeling: Create shared & merged feature pharmacophores using chemical features from three PDB structures Experience level: medium Time needed: 15 minutes
Views
Sequence
Controls
Advanced controls (opt.)
oo Macromolecule view oo Active site view oo Alignment view
oo oo oo oo
PDB File ‘1ke6’ Check & correct ligand Create pharmacophore Add ligand and pharmacophore to alignment view oo Repeat the last 4 steps for ‘1ke7’ and ‘1ke8’ oo Create a ‘shared feature pharmacophore’ from the 3 pharmacophores oo Align the three ligands to the new pharmacophore
oo Ligand box oo Create pharmacophore
oo Set reference element oo Center all structures oo Alignment context menu
(button or menu)
oo Add molecule to align-
ment
oo Add pharmacophore to
alignment
oo Generate shared feature
pharmacophore (by features) oo Generate merged feature pharmacophore
Description: Type ‘1ke6’ in the upper right area of the screen and press the button ‘ ‘1ke6’. The protein will be ed and displayed in macromolecule view. Click on the yellow box to focus on ligand ‘LS2201’. Check the ligand for its correctness. Press ctrl-f9 to create a pharmacophore (cmd-f9 on OS X). Use the data exchange widget and select ‘add molecule to alignment view’ [1a] and the molecule will be added to the alignment view, increasing a counter in the corresponding tab. Use the data exchange widget and select ‘add pharmacophore to alignment view’ [1b] to do the same with the pharmacophore. Repeat these steps with ‘1ke7’, and ‘1ke8’. Finally, you should have six elements in the alignment view. Click on the ‘Alignment’ tab to work with them. Select the three pharmacophores in the alignment list using Ctrl-click or multiple selection [2], and press the ‘Generate shared feature pharmacophore (by features)’ button [3]. A new pharmacophore called ‘Shared [LS2201-1, LS3201-1, LS42-1]’ is appended to the list. Mark this element and make it the reference element using the ‘Set reference’ button, which will change the display style of this element to red and bold. Now use the ctrl key (cmd on OS X) to additionally select the molecules identified by the blue icons [4]. Additionally, hide the three pharmacophores derived from 1ke6, 1ke7, and 1ke8 by clicking on the eye symbols on the right next to these alignment entries [4]. Now press the ‘align’ button. You will see the three ligands aligned to the shared feature pharmacophore derived from the three PDB complexes [5]. Where to go from here: oo Customizing pharmacophore creation preferences oo Virtual screening in LigandScout
1 2
3
4
5
LIGANDSCOUT TUTORIAL CARD 3
Virtual screening and comparing active/inactive hit retrieval rates Experience level: advanced Time needed: 15 minutes
Views
Sequence
Controls
Advanced controls (opt.)
oo Macromolecule view oo Active site view
oo Create pharmacophore
oo Ligand box oo Create pharmacophore
oo Virtual screening with
for virtual screening oo Initiate virtual screening oo Refine pharmacophore model oo Repeat virtual screening
omitted features
(button or menu) oo Initiate virtual screening
Description: Virtual screening aims at the maximum enrichment of active compounds in a hit list. Therefore, such methods are usually validated by assessing the accuracy of discrimination between actives and decoys. A good pharmacophore model will be able to identify a significant portion of known active compounds and as few decoys as possible. In this tutorial, two different types of precalculated multiconformational databases are available: A library containing a selection of known active compounds and a library of decoys. Type ‘1ke7’ in the upper right area of the screen, press the button ‘ 1ke7’ . Create a pharmacophore of the proteinligand complex as described in the LigandScout tutorial sheet 1 ‘Creating a pharmacophore from a single protein-ligand complex‘ [1]. Perform virtual screening using a multi-conformational database with known active compounds (filename ‘cdk2ligands.ldb’) for screening using the ‘Screen pharmacophore against external library’ item located in the ‘Pharmacophore’ menu [2]. Virtual screening will be initiated after selecting the library file (press the butten with the three dots to open a file dialog) and the number of omittable features [3]. Since the initial pharmacophore was derived from all the interactions of only one single ligand, it is rather restrictive and retrieves only few hits. Downsize the number of features by deleting less important features [4] or increase the tolerances of the features by selecting a feature and then choosing ‘Pharmacophore’->’Increase Selected Feature Tolerance’ from the menu. Compare the ratio of active compounds and decoys (library filename ‘cdk2-decoys.ldb’) retrieved by your pharmacophore in order to estimate its predictive power [5]. Further adapt the pharmacophore in order to improve the ratio of actives and decoys obtained by virtual screening: Remove non-essential chemical features and exclusion volume spheres in order to increase sensitivity; adapt the size of the feature spheres by pressing ‘ctrl-shift-plus’ or the corresponding menu entry (‘Pharmacophore - Decrease Selected Feature Tolerance’) or use the pharmacophore from tutorial card 2 using the data exchange widgets [6]. Where to go from here: oo Pharmacophore modeling: Creating shared and merged feature pharmacophores oo Customizing pharmacophore creation preferences oo Analyzing screening results
1
4
2
3
5
6
LIGANDSCOUT TUTORIAL CARD 4
Analyze LigandScout pharmacophore screening results Experience level: medium Time needed: 10 minutes Prerequisites: Virtual screening hit list
Views
Sequence
Controls
Advanced controls (opt.)
oo Library view (spread-
oo Create or import
oo Library view (Spread-
oo Calculation of Gaussion
sheet) oo Hierarchy view
screening results oo Sort virtual screening hit list oo Inspect and compare hits oo Discover controls of the library view
sheet) oo Viewer controls
Shape Similarity Score
Description: Perform virtual screening to obtain a hit list as described on tutorial card 4 ‘Virtual screening and comparing active/inactive hit retrieval rates’ or insert external screening results from a library file via the ‘File - Insert’ menu entry. Results of virtual screening are depicted in the library view [1]. Sort the hit list by a single click on the ‘Pharm. Fit Score’ table header [2]. You may invert the sorting order by an additional click. Select a compound of interest by single click on the corresponding table row to be depicted in the active site view together with the original ligand [3]. Multiple compounds are selected when holding shift or ctrl (cmd on MAC OS X) while clicking. Keep compounds visible independent from the current selection state by clicking the eye symbol in the table left to the molecule name in the library view spreadsheet [4]. The visibility of the core molecule (i.e., the original ligand) and other items in the active site view is toggled with the eye symbol located in the hierarchy view on the right side next to the molecule name [5]. Select a custom color for the core molecule using the square icon next to the eye icon in the hierarchy view. Calculation of the Gaussian Shape Similarity Score offers further functionality for scoring. Use the ‘Calculate Gaussion Shape Similarity Score’ function located in the ‘Library’ menu for calculation. A new property ‘Shape Tanimoto Score’ will be added in the library view and you can use the spreadsheet functionality for sorting and ranking as described above. Where to go from here: oo Pharmacophore modeling: Creating shared and merged feature pharmacophores oo Customizing pharmacophore creation preferences
2
3
4
5
1
LIGANDSCOUT TUTORIAL CARD 5
View and rescore docking poses with pharmacophores Experience level: medium Time needed: 10 minutes Prerequisites: docking poses from external sources Views
Sequence
Controls
oo Library view (spread-
oo oo oo oo oo oo
PDB File Switch to active site view Import docking poses Analyze docking poses Rescore docking poses Align docking poses to pharmacophore oo Discover the viewer controls of the library view
oo Library view (Spread-
sheet) oo Hierarchy view
Advanced controls (opt.)
sheet)
oo Viewer controls
Description: ‘1ke7’ and create the pharmacophore of the protein-ligand complex. Import the corresponding docking poses stored in MDL SDF file format (tutorial filename ‘1ke7-docking.sdf’) via the ’File’ - ‘Insert’ menu entry [1]. Make sure that the ‘Align Molecules Automatically to Shown Pharmacophore’ option located in the ‘Library’ menu is deactivated [2], which ensures that the coordinates of the docking poses relative to the protein are preserved. The compounds are imported to LigandScout including all properties given in the SD file, such as the docking score. Sort the list of docking poses by a single click on the table header corresponding to the docking score [3]. Invert the sorting order by an additional click if necessary. LigandScout offers four different options for rescoring docking poses based on the 3D pharmacophore that is currently visible in LigandScout: The calculation of the pharmacophore fit of imported docking poses considering or ignoring excluded volume spheres, and the calculation of the pharmacophore fit of the docking poses after alignment to the pharmacophore considering or ignoring excluded volume spheres. Alignment to the pharmacophore before scoring may be useful to optimize docking poses [4]. All four modes are accessed via the ‘Library’ menu. You can modify the pharmacophore by deleting features, adding excluded volumes, etc. to reflect the important features of the binding site that should be fulfilled by the highest ranked hits. Alignment of imported docking poses to the shown pharmacophore is performed automatically upon selection of a spreadsheet line (i.e. a single docking pose) if the ‘Align Molecules Automatically to Shown Pharmacophore’ option located in the ‘Library’ menu is activated. Poses may also be aligned to the pharmacophore using LigandScout’s alignment algorithm using the ‘Align’ button located in the upper right corner of the library view [5]. Where to go from here: oo Customizing pharmacophore creation preferences
1
2
4
3
5
5
LIGANDSCOUT TUTORIAL CARD 6
Espresso: Create ligand-based 3D pharmacophores Experience level: intermediate Time needed: 15 minutes
Views
Sequence
Controls
Advanced controls (opt.)
oo Ligand-based Modeling
oo Import ligands oo Cluster ligand-set in 3D oo Generate ligand-based
oo 3D Clustering oo Generate ligand-based
oo Use different settings for
pharmacophore model
pharmacophore model (button or menu)
conformational model generation oo Use different settings for clustering oo Use different settings for ligand-based pharmacophore model generation
Description: Open the smiles file containing your ligands (‘cdk2.smi’) by choosing the ‘File-Open’ menu. The ligands are imported and shown in 2D on the right side of the screen [1]. Set all candidates for the model to ‘training set’ by choosing ‘Ligand-Set -> Flag Selected Molecules as Training Set’. Cluster the ligand set according to their 3D pharmacophore characteristics using the ‘cluster’ button on the bottom of the 3D view [2]. Keep the default setting in the following dialog and start the clustering process by pressing the ‘OK’ button. Espresso will create a new column in the ligand table called ‘Cluster ID’. Sort the table using this column by clicking on the column header ‘Cluster ID’. Set all molecules with cluster ID ‘1’ to test set by using the table controls [3]. Now click on the button ‘Create ligand-based pharmacophore”’[4] using the default values to start pharmacophore generation. A ligand-based model will be created and displayed in the 3D view. Feature patterns [5] in the table indicate, which features are met by which ligand. A click on a colored square indicates, which feature is linked to the corresponding square. The list on the right lower corner of the screen shows several pharmacophore solutions: by selecting a different solution, it is shown in the 3D view with the ligands aligned accordingly [6]. You can use the data exchange controls on the upper right corner of the 3D view to move the currently shown pharmacophore to other views (e.g. screening) or save it using the ‘File -> Save’ menu and selecting a file type that represents pharmacophores. Where to go from here: oo Align structure-based pharmacophore models with ligand-based ones oo Virtual screening in LigandScout
1
2
3
4
5
6
LIGANDSCOUT TUTORIAL CARD 7
Screen several databases using the “Screening Perspective” Experience level: intermediate Time needed: 5 minutes
Views
Sequence
Controls
Advanced controls (opt.)
oo Screening view
oo Add pharmacophore
oo Load Screening Database
oo none
model to screening view oo Load virtual database oo Perform virtual screening
(Button) oo Perform Screening (Button)
Description: In contrast to the virtual screening procedure in the ’Structure-based Modeling’ perspective which is designed for pharmacophore model refinement with respect to ligand-binding pocket properties, the ‘Screening’ perspective provides a fast and convenient way of performing multiple virtual screens with different screening databases against your final pharmacophore model, which can be useful for validation. Import your favorite pharmacophore model (e.g. the one created from tutorial card 1 or 6) into the ‘Screening’ perspective either by using the ‘Exchange Data Widget’ [1] or the ‘Open’ entry in the ‘File’ menu. Click on the ‘Load Screening Database’ button [2] to load the screening database ‘cdk2-ligands.ldb’. Enable the loaded screening database for screening by clicking on the grey button left to the screening database entry [3]. Now, select the pharmacophore model from the ‘Pharmacophores’ list to make sure it is shown in the 3D view. Next to the ‘Load Screening Database’ button, the ‘Perform Screening’ and ‘Delete Pharmacophore’ button will be enabled. Click on the ‘Perform Screening’ button to initiate the virtual screening. View the results in the table at the bottom of the screen [4]. You can add more databases and pharmacophores for further virtual screening experiments. Please make sure that only one pharmacophore is visible before you start screening. Where to go from here: oo Analyze screening results oo Advanced screening: ROC curves and model combination
1
2
4
3
LIGANDSCOUT TUTORIAL CARD 8
Advanced screening: ROC curves and model combination Experience level: advanced Time needed: 15 minutes
Views
Sequence
Controls
Advanced controls (opt.)
oo Screening view
oo Add pharmacophore
oo Load Screening Database
oo Boolean Expression
oo oo oo oo
models to screening view Load virtual database Perform virtual screening Plot ROC curve Combine models by ‘OR’
(Button) oo Perform Screening (Button) oo Plot ROC Curve (Button)
(Field)
Description: In contrast to the virtual screening procedure in the ’Structure-based Modeling’ perspective which is designed for pharmacophore model refinement with respect to ligand-binding pocket properties, the ‘Screening’ perspective provides means for statistical evaluation of screening performance. Import the structure-based pharmacophore models from PDB entry 1ke7 (tutorial card 1) and the ligand-based CDK2 pharmacophore (tutorial card 2) into the ‘Screening’ perspective either by using the ‘Exchange Data Widget’ [1] or, alternatively, load a saved version using the ‘Open’ entry in the ‘File’ menu. Click on the ‘Load Screening Database’ button [2] to load the screening database ‘cdk2-ligands.ldb’, and click the same button again and load the database ‘cdk2-decoys.ldb’. Click once on the tri-state selector left to the database names to mark ‘cdk2-ligands.ldb’ as a database containaing actives (green database symbol), and twice on the tri-state slector left of ‘cdk2-decoys.ldb’ to mark this database as decoys (red database symbol) [3]. Now, select the pharmacophore model ‘1ke7’ from the ‘Pharmacophores’ list to make sure it is shown in the 3D view. Click on the ‘Perform Screening’ button to initiate the virtual screening [4]. After screening has finished, press the ‘Plot ROC Curve’ button [5] to plot a ROC curve [6]. Now select both pharmacophores in the ‘Pharmacophores’ table. Numbers in curly brackets (“{1} and {2}”) indicate the number of the pharmacophore. Enter “1 or 2” into the boolean expression field [7] indicating that a virtual hit in this screening run should either fit into the first (1) or to the second (2) pharmacophore. Press the screening button [4] again to re-screen, and the ROC curve button [5] again to compare the virtual screening statistics. View the virtual hits in the table at the bottom of the screen [8]. You can add more databases and pharmacophores for further virtual screening experiments. Where to go from here: oo Analyze screening results
1
2
3
5
6
7
8
4
Related Documents g842
Ligandscout Tutorial Cards 493i1o
November 2019 13
Cards 5b696f
April 2022 0
October 2019 166
Credit Cards 2261e
December 2021 0
Dixit Cards 33w5u
December 2019 524
Credit Cards 2261e
November 2019 93More Documents from "Irghazi Respayondri" 1k7263
Prosedur Kloning Domba Dolly 53g2x
December 2019 36
Ligandscout Tutorial Cards 493i1o
November 2019 13