Feb 16 2024 Feedback.
Help Videos
Reference Guide
Getting Started
Protein Structure
Molecular Graphics
Slides & ActiveICM
Sequences & Alignments
Protein Modeling
Learn and Predict
Virtual Screening
Molecular Dynamics
3D Ligand Editor
Tables and Plots
Local Databases
 Graphics Tutorials
 3D Documents
 Protein Structure Tutorials
 Sequence and Alignment Tutorials
 Homology Modeling Tutorials
 Cheminformatics Tutorials
 3D Ligand Editor Tutorial
 APF Tutorial
 Ligand Docking Tutorials
 Protein-Protein Docking
 Virtual Ligand Screening Tutorials
 Induced Fit (Flexibility) Docking Tutorials
 2D QSAR Model
 3D QSAR Tutorial
PrevICM User's Guide
21.13 Chemical FingerPrint QSAR Model Tutorial

| Documentation |

In this example we build a 2D QSAR FingerPrint model based on pKd data for a selection of inhibitors to a nuclear hormone receptor Thyroid Hormone Receptor Alpha using 2D Chemical FingerPrints.

Note. It is usually recommended to Convert the IC50 to pIC50 (-LogIC50)

Step 1: Download and open the prepared ICM file. The file can be downloaded here https://molsoft.com/addons/ftp/workshop (1. cut and paste the address into a web browser. 2. Right click on qsar_2D_example.icb and choose "Save Link As". 3. Open the .icb in ICM File/Open). Notice the file contains two tables: we will use THA_TrainingSet to build the QSAR model and THA_External_Test_Set to test the model. Each table contains experimental pKd from ChEMBL.

Step 2: Build the model using the options in Chemistry/Build Prediction Model. Use the options shown in the picture above - each option is described in more details here. In this example we will use the PLS regression algorithm to build the model.

Step 3: Once the model is built a table displaying the model properties will be displayed. The self and test R2 values are shown giving an indication of the quality of the model. The model is saved in the ICM workspace - right clicking on the model in the ICM workspace allows you to view the model data.

Step 4: One way to visualize the quality of the model is to plot the experimental pKd (pKd) versus the predicted pKd (pKdpred). There are other options in the right click menu to improve the model or look at the`model-weight{weights}.

Step 5: When you are happy with your model you can use it to predict pKd for the external test set in the THA_External_Test_Set table. Use the options in Chemistry/Predict.

Step 6: A new column containing the predicted pKd will be displayed.

Step 7: Plot experimental pKd (pKd) versus Predicted pKd (pKdpred)to visualize the data and observe reported R2 value.

Explicit Group
3D QSAR Tutorial

Copyright© 1989-2020, Molsoft,LLC - All Rights Reserved.
This document contains proprietary and confidential information of Molsoft, LLC.
The content of this document may not be disclosed to third parties, copied or duplicated in any form,
in whole or in part, without the prior written permission from Molsoft, LLC.