202,716 De Novo Designed Small Molecules Targeting Mpox Virus VP39 Protein

202,716 unique small molecules designed to target the Mpox virus VP39 methyltransferase protein. The dataset was generated using deep reinforcement learning and published by Irina Tirosyan on figshare in April 2026. It includes annotations for docking scores, physicochemical properties, ADMET predictions, and synthetic accessibility.

Use Cases

• Virtual screening for Mpox antiviral candidates based on docking scores and drug-likeness filters.
• Lead optimization by analyzing physicochemical descriptors and structural alerts.
• Prioritizing compounds for experimental validation using predicted ADMET properties.
• Assessing synthetic feasibility for candidate molecules based on retrosynthetic pathway information.

Strengths

• Contains 202,716 unique, de novo designed compounds.
• Provides extensive annotations per molecule, including docking scores, ADMET predictions, and drug-likeness compliance.
• Includes information on commercial availability and retrosynthetic feasibility.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

figshare, author Irina Tirosyan

Collection Method

Generated using a deep reinforcement learning approach optimized for binding affinity, drug-likeness, and synthetic accessibility.

Time Range

Dataset publication date is 2026.

Freshness

Last updated 2026-04-13 07:48:35

Geography

Global (computational/therapeutic target).