AI-Prioritized Enzyme Variants for Biocatalytic Amine Synthesis

A dataset from 2026 describes an AI-driven protein-engineering framework for biocatalyst development. It likely contains structural and energetic descriptors for enzyme variants, derived from a library of 1.39 million structural fragments. The data was authored by Pravin Kumar R and shared on figshare.

Use Cases

• Training machine learning models for enzyme property prediction based on structural fragment descriptors.
• Benchmarking computational protein design methods against experimental validation data for solubility and catalytic efficiency.
• Analyzing the relationship between mutation combinations and industrial process metrics like conversion rate and enantiomeric excess.

Strengths

• The underlying method was validated at a 7 L fermentation scale, suggesting industrially relevant data.
• The AI framework integrated interaction energy descriptors and solvent effects, indicating a multi-faceted feature set.
• The dataset is derived from a library of 1.39 million structural fragments, implying a substantial foundation for feature generation.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small (22.7 KB), indicating limited scope, likely containing summary results or feature descriptors rather than raw screening data.

Provenance

Source

figshare

Collection Method

Likely generated through computational protein design and AI-driven variant prioritization, followed by experimental validation.

Freshness

Last updated 2026-05-14 12:34:41; freshness should be verified.