DrugLikenZ: An Open-Source Graphical User Interface for Multi-Rule Drug-Likeness and Batch Property Evaluation

Mahdaoui, Y., & Zakkoumi, H. (2026)

Version 1.0.2 | 91.9 MB via SourceForge

---

Overview

DrugLikenZ is a Python-based graphical user interface (GUI) designed for the rapid evaluation of chemical libraries against established multi-rule drug-likeness criteria. The application automates property calculation and compliance checking across multiple filters (Lipinski, Veber, Muegge, and RO3), providing researchers with an intuitive visual heatmap of their results. By integrating local RDKit calculations with optional PubChem database queries, it offers a high-throughput workflow for preliminary drug discovery and chemical education.

Technical Abstract

The core concept of this application is to rapidly predict the drug-likeness of chemical compounds by applying multiple established physicochemical filters, including Lipinski's Rule of Five (Ro5), Veber’s Rule, Muegge’s Method, and the Rule of Three (RO3). By evaluating parameters such as molecular weight, lipophilicity, flexibility, and surface area, the tool provides a comprehensive assessment of oral bioavailability, permeability, and lead-like potential for drug discovery.

Calculation Method

The calculation method involves using RDKit and querying the PubChem database via the compound SMILES string to retrieve or calculate key molecular properties:

• MW: Molecular Weight
• HBA: Hydrogen Bond Acceptor count
• HBD: Hydrogen Bond Donor count
• LogP: Lipophilicity (Partition Coefficient)
• ROTB: Rotatable Bond count (Molecular Flexibility)
• PSA: Polar Surface Area
• Rings: Number of cyclic structures
• Carbons: Total Carbon atom count
• Heteroatoms: Count of non-carbon/non-hydrogen atoms

Parameter Definitions

The drug-likeness of a molecule is evaluated using several physiochemical properties: Molecular Weight (MW) measures the total mass of atoms, influencing a drug's ability to cross biological membranes. Hydrogen Bond Acceptors (HBA) and Donors (HBD) count atoms capable of forming hydrogen bonds, which are critical for binding affinity and solubility. LogP (Partition Coefficient) determines lipophilicity, indicating the balance between water and fat solubility. Rotatable Bonds (ROTB) quantify molecular flexibility, where fewer bonds often correlate with better oral bioavailability. Polar Surface Area (PSA) calculates the surface sum of polar atoms to predict membrane permeability. Rings, Carbons, and Heteroatoms (non-C/H atoms like N or O) are used to quantify the structural complexity and chemical diversity of the compound.

The application performs a binary evaluation (Pass/Fail) against specific thresholds. A compound is considered drug-like (accepted) if it violates no more than one of these four rules, meaning its total compliance score must be 3 or 4.

Rule Comparison Summary

This table compares the four screening methods available in DrugLikenZ. While Lipinski focuses on size and bonding, the Veber rule specifically targets flexibility (ROTB) and surface area (PSA).

Parameter	Lipinski's Rule of Five	RO3 (Congreve et al.)	Muegge Method	Veber Rule
MW	<= 500	< 300	200 to 600	—
LogP	<= 5	<= 3	-2 to 5	—
HBA	<= 10	<= 3	<= 10	—
HBD	<= 5	<= 3	<= 5	—
ROTB	—	<= 3	<= 15	<= 10
PSA	—	<= 60	<= 150	<= 140
Rings	—	—	<= 7	—
Carbons	—	—	> 4	—
Heteroatoms	—	—	> 1	—
Acceptance	Min. 3/4 pass	All 6/6 pass	All 9/9 pass	All 2/2 pass

Key Features

• Multiple Rulesets: Supports Lipinski's Rule of Five, Miles Congreve et al. Rule of Three (RO3), Muegge Method, and Veber et al. rules.
• Automated Data Retrieval: Interfaces with the PubChem PUG REST API for compound metadata.
• Visual Analytics: Generates dynamic, color-coded heatmaps where green represents compliance and red represents violations.
• Intelligent Formatting: Names of compounds failing the selected criteria are automatically highlighted in red on the axis.
• Large Dataset Management: Includes a navigation system to handle files with hundreds of compounds.
• Clean Export: Save accepted candidates directly to a CSV file and export high-resolution heatmaps as PNG images.

Requirements

The application requires Python 3.x and the following dependencies:

• customtkinter
• rdkit
• pandas
• seaborn
• matplotlib
• pubchempy
• requests

Installation

1. Clone the repository: bash git clone [https://github.com/yourusername/DrugLikenZ.git](https://github.com/yourusername/DrugLikenZ.git)

2. Install the necessary packages: bash pip install customtkinter rdkit pandas seaborn matplotlib pubchempy requests

3. Run the application: bash python main.py

Usage

1. Launch the application.
2. Select your desired filtering rule from the dropdown menu.
3. Use the Browse button to upload a CSV or TSV file containing SMILES strings.
4. Specify the name of the SMILES column in the entry box.
5. The application will automatically deduplicate and process the list.
6. Use the navigation buttons to browse the compliance heatmaps.
7. Click Export to save results or the heatmap image.

Citation

If you use this software in your research, please use the following citation:

Mahdaoui, Y., & Zakkoumi, H. (2026). DrugLikenZ: An Open-Source Graphical User Interface for Multi-Rule Drug-Likeness and Batch Property Evaluation. Zenodo. https://doi.org/10.5281/zenodo.18500549

License

This project is licensed under the Creative Commons Attribution 4.0 International. The Creative Commons Attribution license allows re-distribution and re-use of a licensed work on the condition that the creator is appropriately credited.