Bayexer

Bayexer performs Bayesian demultiplexing of Illumina sequencing reads to assign reads to their originating samples using index sequences.

Key Features:

• Bayesian demultiplexing: Uses a Bayesian framework to assign sequencing reads to samples based on index sequences.
• Contaminant-based training: Leverages contaminant sequences found within target reads as training data for classification.
• Naïve Bayes classifier: Applies a naïve Bayes classifier to compute assignment probabilities for reads.
• Low-quality dataset handling: Enhances assignment accuracy in datasets with low-quality index reads.
• Alternative to similarity metrics: Avoids sole reliance on similarity metrics between read indices and reference index sequences.
• Illumina compatibility: Designed for sequencing reads generated by Illumina sequencers.
• Implementation: Implemented in Perl.
• Performance: Reported to provide improved assignment accuracy and speed relative to existing demultiplexing approaches.

Scientific Applications:

• Genomics experiments: Assigns reads to samples in Illumina-based genomics studies.
• Contaminated sequencing datasets: Enables reliable read assignment when target reads contain contaminant sequences.
• Low-quality sequencing runs: Applied to demultiplexing of datasets with degraded or low-quality index information.

Methodology:

Bayesian demultiplexing using index-sequence information, leveraging contaminant sequences within target reads as training data for a naïve Bayes classifier; implemented in Perl.