LRCstats

LRCstats evaluates the accuracy of long-read correction methods for third-generation sequencing (TGS) data from PacBio and Oxford Nanopore platforms.

Key Features:

• Accuracy assessment: Measures correction accuracy for noisy long reads produced by TGS platforms.
• Reference-free evaluation: Computes accuracy without requiring mapping corrected reads onto a reference genome.
• Simulator-based ground truth: Leverages long reads simulators that provide each simulated read with an alignment to its originating reference genome segment.
• Error-profile consistency: Aligns accuracy measurements with the error profiles introduced during the simulation process.
• Platform specificity: Applicable to reads from PacBio and Oxford Nanopore sequencing technologies.
• Benchmarking demonstration: Demonstrated by analyzing four hybrid correction methods for PacBio long reads across three datasets.

Scientific Applications:

• Benchmarking correction tools: Quantitatively compare performance of long-read correction methods for TGS data.
• Hybrid method evaluation: Evaluate and compare hybrid correction methods for PacBio long reads.
• Dataset-specific assessment: Assess correction performance across multiple datasets.
• Simulation-aware accuracy studies: Study the impact of simulated error profiles on correction accuracy.

Methodology:

Uses long reads simulators that attach each simulated read to an alignment of its originating reference genome segment to measure correction accuracy without mapping corrected reads to a reference, and aligns accuracy measurements with the error profiles introduced during simulation; demonstrated on four hybrid correction methods for PacBio across three datasets.