BrEPS 2.0: Optimization of sequence pattern prediction for enzyme annotation

Area of Science:

• Genomics and Bioinformatics
• Enzymology
• Computational Biology

Background:

• High-throughput sequencing generates vast datasets, overwhelming manual gene function annotation.
• Automatic sequence pattern-based methods like BrEPS are needed for efficient gene function prediction.
• Previous BrEPS versions showed promise but required optimization for current large-scale data.

Purpose of the Study:

• To reimplement and optimize the BrEPS protocol for handling larger datasets efficiently.
• To enhance the accuracy and scope of enzymatic function prediction.
• To improve the usability and accessibility of the BrEPS database and tools.

Main Methods:

• Enhanced data selection using Swiss-Prot and TrEMBL sequences.
• Introduction of extended patterns incorporating semi-conserved amino acid positions.
• Integration of consensus EC numbers and IUBMB enzyme nomenclature for function prediction.

Main Results:

• Optimized BrEPS achieves reliable enzyme function prediction on large datasets within acceptable timescales.
• Extended patterns increase matching capabilities while preserving essential structural information.
• The Braunschweig Enzyme Database (BRENDA) now features a redesigned website and downloadable tools.

Conclusions:

• The optimized BrEPS protocol is a scalable and efficient tool for predicting gene enzymatic functions.
• Enhanced pattern generation and data selection improve prediction accuracy and broaden applicability.
• Freely accessible website and command-line tools facilitate large-scale sequence analysis and research.