PEAR

PEAR merges Illumina paired-end reads by evaluating all possible overlaps to produce accurate extended sequences for downstream genomic analyses.

Key Features:

• Versatility in Fragment Lengths: Does not require prior knowledge of the target fragment size, enabling merging across variable fragment lengths.
• Comprehensive Overlap Evaluation: Evaluates all possible overlaps between paired-end reads to identify optimal merge positions.
• Statistical Validation: Applies a statistical test/framework to minimize false-positive merges.
• Performance Efficiency: Produces highly accurate merged reads on simulated and empirical datasets and can process millions of paired-end reads in minutes on standard desktop hardware.
• Parallel Processing Capability: Offers a parallel version that achieves linear speedups compared to the sequential implementation on multi-core architectures.
• Implementation: Implemented in C and uses POSIX threads for parallel execution.

Scientific Applications:

• Read length extension for genomic analyses: Extends Illumina read lengths through merging to support analyses that benefit from longer contiguous sequences, including detection of genetic sequences and structural variations.
• High-throughput variable-fragment datasets: Applicable to processing Illumina high-throughput sequencing datasets with variable fragment sizes that challenge conventional merging algorithms.

Methodology:

Assesses all feasible overlaps between paired-end reads without requiring prior knowledge of the target fragment size and applies a statistical test to reduce false-positive merges.