TagDust2

TagDust2 extracts mappable reads from raw next-generation sequencing (NGS) data by modeling complex read architectures with hidden Markov models to remove barcodes, adaptors, contaminants, and low-complexity sequences for downstream analyses.

Key Features:

• Hidden Markov Models (HMM): Implements a library of HMMs to model and identify complex patterns in read architectures for precise read extraction.
• Support for Multiplexed Data: Processes multiplexed datasets including single-end and paired-end libraries and reads containing unique molecular identifiers (UMIs).
• Post-Processing Capabilities: Performs post-processing to exclude known contaminants and filter out low-complexity sequences from extracted reads.
• Automatic Library Type Detection: Detects library type automatically from a predefined selection to classify read architectures prior to extraction.
• Integrated Read-to-Mappable Conversion: Integrates extraction, contaminant exclusion, and filtering steps to convert raw NGS reads into mappable sequences.

Scientific Applications:

• Variant Calling: Produces high-quality mappable reads suitable for accurate variant calling analyses.
• Transcriptome Profiling: Extracts reads appropriate for transcriptome profiling workflows by removing adaptors, barcodes, and artefacts.
• Metagenomic Studies: Cleans and filters reads from complex samples to improve downstream metagenomic analyses.

Methodology:

Uses a library of hidden Markov models to model read architectures, supports single-end, paired-end and UMI-containing multiplexed data, performs contaminant exclusion and low-complexity filtering as post-processing, and detects library type from a predefined selection.