Genome Classifier

Genome Classifier classifies genomic sequences by predicting their genome of origin using a naïve Bayesian classifier implemented in Python/pyrex, leveraging nucleotide composition and oligonucleotide frequency profiles such as GC content.

Key Features:

• Bayesian Classification: Uses a naïve Bayesian approach to predict the genome of origin for specific genomic sequences.
• Implementation: Implemented in Python/pyrex.
• Reference Scope: Analyzes complete sequences from a diverse set of eubacterial and archaeal genomes as reference data.
• Genomic Signature Analysis: Exploits nonrandom sequence variations in short oligonucleotide frequency profiles and GC content as distinguishing features.
• Short-Sequence Performance: Capable of classifying sequences as short as 400 bases with reported accuracy up to 85%.

Scientific Applications:

• Biodiversity Studies: Supports microbial diversity and evolutionary biology analyses by assigning short sequences to probable source genomes.
• Horizontal Gene Transfer Detection: Applied to identify horizontal gene transfer events, including prediction of transfers such as sodC and bioC from Haemophilus influenzae to Neisseria meningitidis with donor and recipient species identified.

Methodology:

Implements a naïve Bayesian classifier in Python/pyrex; computes GC content and frequency profiles of specific short oligonucleotides from complete eubacterial and archaeal genomes; applies these profiles to predict genome of origin for input sequences and to infer potential horizontal gene transfer events.