DeepECtransformer: Predicted Enzyme Functions for Protein Sequences

Supplementary datasets support the paper "Functional annotation of enzyme encoding genes using deep learning with transformer layers". The collection includes predicted Enzyme Commission numbers for sequences from Swiss-Prot and NCBI, along with visualizations and motif analyses. Author Gi Bae Kim released this data under an Open Access license.

Use Cases

• Benchmarking enzyme function prediction models based on the provided EC number predictions for Swiss-Prot sequences.
• Analyzing conserved sequence motifs for specific enzyme functions using the highlighted motifs data.
• Studying genomic enzyme potential across species based on EC number predictions for over 128 million sequences from NCBI genomes.
• Investigating protein solubility as a downstream property using the solubility prediction results for y-ome proteins.

Strengths

• Includes predictions for a large scale of 128,100,490 amino acid sequences from 70,600 genomes in NCBI.
• Provides multiple complementary data types, including predictions, visualizations (TMAP), and motif analyses.
• Source data is linked to a peer-reviewed scientific paper, providing methodological context.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for individual supplementary datasets is unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

Gi Bae Kim, associated with the paper published on paperswithcode.

Collection Method

Data generated through the DeepECtransformer deep learning model for functional annotation.