UrQt

UrQt performs unsupervised quality trimming of Next Generation Sequencing (NGS) reads using a probabilistic segmentation model to maximize retention of informative nucleotides for downstream analyses.

Key Features:

• Unsupervised quality trimming: Implements a probabilistic framework to autonomously determine the optimal segmentation between unreliable and informative nucleotide segments within reads.
• Minimal quality threshold (Phred score): Requires a single user-specified minimal quality threshold (Phred score) to classify nucleotides as informative versus unreliable.
• C++ implementation and parallelization: Implemented in C++ with a parallelized architecture and low memory footprint for rapid processing of large-scale genomic datasets.
• Empirical performance: Demonstrated an optimal tradeoff between number of trimmed nucleotides and quality objective across seven RNA and DNA sequencing experiments, yielding increased quantity and quality of retained reads.

Scientific Applications:

• NGS data preprocessing: Produces trimmed read sets that preserve informative nucleotides for general preprocessing of sequencing datasets.
• Variant calling: Provides a higher quantity and quality of retained reads that support downstream variant detection workflows.
• Transcriptome analysis: Retains informative bases in RNA-seq reads to improve data available for transcriptome quantification and analysis.
• Metagenomics: Preserves informative nucleotides in metagenomic reads to enhance community profiling and downstream analyses.

Methodology:

Detects the best segmentation between unreliable and reliable nucleotide segments using a probabilistic model and trims bases below the user-specified minimal Phred quality threshold.