The Materials Informatics Toolkit is a powerful, user-friendly JMP add-in designed to provide an interface for using the RDKit package to calculate descriptors from Simplified Molecular Input Line Entry System (SMILES). I developed the add-in after numerous conversations with JMP users who were working with molecular structures and using Python for their analyses. They needed to be able to perform complex analyses – but without extensive coding.
The Materials Informatics Toolkit offers a point-and-click option for anyone who wants to use JMP to handle chemical structure data. It's helpful in a variety of instances, from searching for new molecular candidates that meet specific requirements to analyzing relationships between molecular structures and physical properties.
Before getting started, please remember that this add-in assumes that you already have SMILES data to analyze. SMILES represents molecular structures using strings.
How to use the Materials Informatics Toolkit
After downloading and installing the add-in:
- In JMP, open the Add-Ins menu in the top navigation and click Toolkit for Materials Informatics.
- Click Descriptor Calculation.
- Select the column containing SMILES.
- Select the features you want to generate.
- Click OK to send the data from JMP to Python. Python generates features and sends the data back to JMP. Creating a model with these features allows you to predict the characteristic values of a certain substance.
- To draw a picture from this molecular information, go back to the add-in menu and click Drawing Molecules.
- Select the SMILES column and click OK. This sends the data from JMP to Python, creates a Python diagram, and sends the data back to JMP. The resulting images are stored in a column in the table.
Watch the demo
This video demonstrates how to use the Materials Informatics Toolkit:
Here are a few helpful hints as you use the Materials Informatics Toolkit:
- This add-in uses a Python integration feature. RDKit must be installed before using this add-in.
- Once you have created the descriptors using this add-in, you can use this information to build the model to predict molecular properties and so on. If you have JMP Pro, you can use Model Screening and/or another platform as usual (you can also build the model using JMP).
- This add-in can create a molecular image using SMILES. You can create a new column that contains images and then use that as your label column.
The JMP Marketplace has add-ins that use Python integration that open doors to new possibilities. Please give this and other add-ins a try and then share any ideas you have for improvements with us.