The B6 Database (B6DB)

The B6 Database (B6DB) catalogs and classifies pyridoxal 5'-phosphate (PLP)-dependent enzymatic activities and their protein families to support identification and comparative analysis of PLP-dependent enzymes.

Key Features:

• Repository of PLP-dependent activities and families: Integrates information linking PLP-dependent enzymatic activities to corresponding protein families.
• Family definition: Defines protein families as monophyletic groups of sequences that share the same enzymatic function.
• Activity-to-family mapping: Associates one or more protein families with each of 121 documented PLP-dependent activities for which sequence data are available.
• HMM-based classification: Uses Hidden Markov Models (HMMs) built from alignments of protein family sequences to facilitate functional classification.
• Genomic dataset analysis: Applies HMMs to predicted protein sequences in genomic datasets to identify and classify PLP-dependent enzymes.
• Gene census capability: Enables censuses of human genes coding for PLP-dependent enzymes.
• Comparative analyses support: Supports identification, classification, and comparative studies of B6-dependent enzymes and their evolutionary relationships.

Scientific Applications:

• Functional annotation of genomes: Annotating PLP-dependent enzymatic functions in predicted protein sets from genomic datasets using HMMs.
• Human gene censuses: Enumerating and classifying human genes that encode PLP-dependent enzymes.
• Comparative and evolutionary analysis: Comparing protein families and activities to study divergence and evolutionary relationships among PLP-dependent enzymes.
• Activity-family association studies: Associating specific enzymatic activities with monophyletic protein families to link function and sequence lineage.

Methodology:

Protein families are defined as monophyletic groups of sequences sharing the same function; alignments of family sequences are used to build Hidden Markov Models (HMMs), which are applied to predicted protein sequences in genomic datasets for functional classification.