Trowel

Trowel performs error correction of Illumina sequencing reads by identifying trusted k-mers via a k-mer spectrum quality-threshold approach to improve base-call accuracy and base quality scores.

Key Features:

• Massively Parallelized Architecture: Implements massively parallel processing to scale error correction to large Illumina datasets.
• k-mer Spectrum with Quality Threshold: Uses a k-mer spectrum approach that selects trusted k-mers based on a quality threshold rather than coverage cutoffs.
• Error Correction and Quality Enhancement: Corrects erroneous base calls and improves base quality scores by leveraging trusted k-mers.
• Adaptability Across Diverse Conditions: Performs robustly across varying coverage levels, genome sizes, read lengths, and fragment sizes.
• Efficient Data Structures: Employs efficient data structures to reduce data-access latency and accelerate k-mer analysis.
• Implementation: Implemented in C++.

Scientific Applications:

• Short-read sequencing accuracy: Improves accuracy of Illumina short-read datasets for downstream analyses.
• Genome assembly improvement: Reduces assembly artifacts by minimizing erroneous base calls that can lead to misassemblies.
• RNA editing inference: Reduces false inferences of RNA editing by correcting sequencing errors that mimic editing events.
• Uneven coverage datasets: Handles uneven coverage effectively to produce more reliable datasets for downstream analyses.

Methodology:

Analyzes k-mer frequency and distribution within sequencing reads using a k-mer spectrum approach, selects trusted k-mers by applying a quality threshold instead of coverage cutoffs, and uses those trusted k-mers to correct base-call errors and improve base quality scores.