EBI patent sequence database

EBI patent sequence database provides non-redundant, annotated nucleotide and protein sequences extracted from patent documents to support sequence retrieval, clustering, and patent-specific analyses.

Key Features:

• Non-Redundant Databases: Aggregates non-redundant sequence sets covering the EMBL-Bank nucleotides patent class and patent protein databases by eliminating duplicate sequences.
• Value-Added Annotations: Incorporates annotations derived from patent documents including publication number corrections, earliest publication dates, and feature collations.
• Hierarchical Clustering: Implements two-level clustering based on MD5 checksums with Level-1 clusters grouping sequences that are 100% identical over their entire length and Level-2 sub-clustering using patent family information.
• Comprehensive Coverage: Includes both nucleotide sequences from the EMBL-Bank patent class and patent-associated protein sequences.

Scientific Applications:

• Enhanced Data Retrieval: Enables precise identification and retrieval of specific sequences and their patent metadata by reducing redundancy and improving annotation quality.
• Cross-Disciplinary Research: Provides access to biological sequences and annotations embedded in patent documents for bioinformatics and interdisciplinary studies.
• Intellectual Property Analysis: Supports patent validity and scope assessments through MD5-based clustering and patent-derived annotations.

Methodology:

Clustering sequences using MD5 checksums into Level-1 (100% identity) and Level-2 (patent-family based) clusters and incorporating annotations extracted from patent documents such as publication number corrections, earliest publication dates, and feature collations.