πŸ”¬ Workflows & Demonstrations


πŸ–₯️ The User Interface

ADMETShiny (Version 1.0.0) is an open-source R Shiny web application designed for the management, calculation, filtering, and selection of drug-like molecules in early-stage drug discovery.

Landing page panels

Below the navigation bar, the dashboard is split into two sections:

NoteπŸ“ Description

Summarizes the app’s integration of bioinformatics and cheminformatics workflows for molecular descriptor visualization.

TipπŸ“– How to cite?

Provides the formal 2026 academic citation and the GitHub repository link.


πŸ§ͺ The Dataset Example

The Drug-likeness Filters section lets users apply customizable filters based on popular medicinal chemistry criteria:

  • βœ… Lipinski
  • βœ… Veber
  • βœ… Ghose
  • βœ… Egan
  • βœ… Muegge

Adjustable thresholds are available for maximum molecular weight, logP, hydrogen bond acceptors/donors, and maximum Lipinski violations.

Filtered results

The central area displays the Filtered Results table under the Filtered Molecules tab (6), giving a structured overview of the compounds that meet the selected criteria. This interactive data grid shows key molecular descriptors:

  • Molecule ID
  • Canonical SMILES
  • Chemical formula
  • Molecular weight (MW)
  • Number of heavy atoms
  • Number of aromatic heavy atoms
TipπŸ” Quick search

Use the Search bar (8) at the top right of the table to filter molecules by name, SMILES string, or specific properties.

The top tabs, Data Preview (5) and Plots (7), let you toggle seamlessly between the raw dataset, refined subsets, and data distributions, providing an intuitive, integrated workflow for compound prioritization.


πŸ“Š You Probably Need a Report

The executive summary and statistical breakdown confirm that 100% of the library was successfully retained.

Key metrics

Metric Value
🧬 Molecular Weight (mean) 392.44
πŸ’§ LogP (mean) 3.33
πŸ”· TPSA (mean) 90.53
⭐ Drug-likeness score (mean) 85.4 / 100
πŸ† Compounds rated β€œExcellent” 80.5%
ImportantHigh-quality library

The majority of compounds cluster within the highest drug-likeness tier, confirming the overall quality of the library.

Supporting visualizations

Maps gastrointestinal absorption and blood-brain barrier permeability, alongside P-glycoprotein efflux status.

A stacked bar chart detailing individual rule violations across the standard medicinal chemistry criteria: Lipinski, Ghose, Veber, Egan, and Muegge.

A histogram that visually confirms the sharp clustering of most compounds within the highest drug-likeness tier.

Back to top