Uploaded November 2025 | Updated September 2026, 3 weeks ago
Learn how to bridge cheminformatics and pharmacokinetics with this complete, GPU-accelerated drug discovery workflow in MATLAB®.
This demonstration builds on concepts from another example, Visualize and Analyze Molecular Structures, and moves into a practical cheminformatics pipeline aimed at identifying potential SGLT2 inhibitors for treating type 2 diabetes. In this example, a large molecular data set is used to call RDKit and accelerate molecular fingerprint calculations with GPU capabilities in MATLAB. This enables large-scale Tanimoto similarity analysis to quickly identify compounds structurally similar to known FDA-approved inhibitors. Molecules above a similarity threshold are saved for further analysis.
Next, these screening results are connected to pharmacokinetics using SimBiology®. Specifically, how these key molecular descriptors—LogP (lipophilicity), LogS (aqueous solubility), and molecular weight—affect bioavailability, a critical factor in determining whether an orally administered drug reaches the bloodstream in active form, is explored. The model simulates the dissolution, absorption, and distribution of each candidate using a compartmental approach (solid drug in the GI tract, dissolved drug, and plasma). Permeability is modeled as a function of LogP and molecular weight.
High-accuracy ODE simulations are then configured and run to predict plasma concentration profiles over 12 hours for each candidate molecule. The results help visualize how structural and physicochemical properties influence in vivo drug performance—turning chemical structure data into actionable pharmacological insights.
Finally, this work is extended by integrating SGLT2 inhibition into a physiologically-based glucose-insulin model, exploring how absorption properties might impact therapeutic outcomes at the whole-body level.
This workflow demonstrates:
- Data-driven screening of drug candidates using cheminformatics and GPU acceleration
- Mechanistic PK modeling to predict bioavailability from molecular properties
- A clear connection between molecular descriptors and system-level drug action
Watch now to see how you can combine similarity search, molecular visualization, and bioavailability modeling in one seamless workflow, powered by MATLAB, RDKit, and SimBiology.
Learn more:
- Watch our previous video on visualizing molecular structure: youtube.com/watch?v=URQcL6N996E
- Learn more about MATLAB's Simbiology Toolbox: bit.ly/4ocXFjT
- Download and try the MATLAB code: bit.ly/46HG4e1
--------------------------------------------------------------------------------------------------------
Get a free product trial: goo.gl/ZHFb5u
Learn more about MATLAB: goo.gl/8QV7ZZ
Learn more about Simulink: goo.gl/nqnbLe
See what's new in MATLAB and Simulink: goo.gl/pgGtod
© 2025 The MathWorks, Inc. MATLAB and Simulink are registered trademarks of The MathWorks, Inc.
See mathworks.com/trademarks for a list of additional trademarks. Other product or brand names may be trademarks or registered trademarks of their respective holders.
Learn how to bridge cheminformatics and pharmacokinetics with this complete, GPU-accelerated drug discovery workflow in MATLAB®.
This demonstration builds on concepts from another example, Visualize and Analyze Molecular Structures, and moves into a practical cheminformatics pipeline aimed at identifying potential SGLT2 inhibitors for treating type 2 diabetes. In this example, a large molecular data set is used to call RDKit and accelerate molecular fingerprint calculations with GPU capabilities in MATLAB. This enables large-scale Tanimoto similarity analysis to quickly identify compounds structurally similar to known FDA-approved inhibitors. Molecules above a similarity threshold are saved for further analysis.
Next, these screening results are connected to pharmacokinetics using SimBiology®. Specifically, how these key molecular descriptors—LogP (lipophilicity), LogS (aqueous solubility), and molecular weight—affect bioavailability, a critical factor in determining whether an orally administered drug reaches the bloodstream in active form, is explored. The model simulates the dissolution, absorption, and distribution of each candidate using a compartmental approach (solid drug in the GI tract, dissolved drug, and plasma). Permeability is modeled as a function of LogP and molecular weight.
High-accuracy ODE simulations are then configured and run to predict plasma concentration profiles over 12 hours for each candidate molecule. The results help visualize how structural and physicochemical properties influence in vivo drug performance—turning chemical structure data into actionable pharmacological insights.
Finally, this work is extended by integrating SGLT2 inhibition into a physiologically-based glucose-insulin model, exploring how absorption properties might impact therapeutic outcomes at the whole-body level.
This workflow demonstrates:
- Data-driven screening of drug candidates using cheminformatics and GPU acceleration
- Mechanistic PK modeling to predict bioavailability from molecular properties
- A clear connection between molecular descriptors and system-level drug action
Watch now to see how you can combine similarity search, molecular visualization, and bioavailability modeling in one seamless workflow, powered by MATLAB, RDKit, and SimBiology.
Learn more:
- Watch our previous video on visualizing molecular structure: youtube.com/watch?v=URQcL6N996E
- Learn more about MATLAB's Simbiology Toolbox: bit.ly/4ocXFjT
- Download and try the MATLAB code: bit.ly/46HG4e1
--------------------------------------------------------------------------------------------------------
Get a free product trial: goo.gl/ZHFb5u
Learn more about MATLAB: goo.gl/8QV7ZZ
Learn more about Simulink: goo.gl/nqnbLe
See what's new in MATLAB and Simulink: goo.gl/pgGtod
© 2025 The MathWorks, Inc. MATLAB and Simulink are registered trademarks of The MathWorks, Inc.
See mathworks.com/trademarks for a list of additional trademarks. Other product or brand names may be trademarks or registered trademarks of their respective holders.
