Be-dataHIVE: a base editing database.

Area of Science:

• Biotechnology
• Genetics
• Bioinformatics

Background:

• Base editing offers precise single nucleotide transformations for treating rare diseases.
• Current base editor design is labor-intensive, with unpredictable outcomes and limitations in accuracy and efficiency.
• Minimizing bystander mutations is critical for clinical applications of base editing.

Purpose of the Study:

• To develop a comprehensive and accessible dataset for improving computational models of base editing.
• To create BE-dataHIVE, a mySQL database containing extensive gRNA target combinations and relevant features.
• To facilitate the advancement of accurate and efficient base editing technologies.

Main Methods:

• Compiled data from five studies to create a large-scale dataset.
• Integrated melting temperatures and energy terms to enrich the data.
• Computed and structured multiple data formats suitable for machine learning applications.
• Developed a mySQL database (BE-dataHIVE) with over 460,000 gRNA target combinations.

Main Results:

• BE-dataHIVE provides a robust dataset exceeding 460,000 gRNA target combinations.
• The database includes enriched features like melting temperatures and energy terms.
• Multiple machine-learning-ready data structures are pre-computed and available.
• The database is accessible via a website and API.

Conclusions:

• BE-dataHIVE addresses the critical need for large, feature-rich datasets in base editing research.
• The database enhances the development and performance of computational models for predicting base editing outcomes.
• BE-dataHIVE supports both practitioners and machine learning researchers in advancing gene editing technologies.