ngsComposer

ngsComposer performs empirical quality filtering of next-generation sequencing (NGS) reads by leveraging known sequence motifs to estimate per-base error rates and remove erroneous base calls and adapter contamination.

Key Features:

• Empirical motif-based error estimation: Leverages known sequence motifs to empirically estimate per-base error rates.
• Erroneous base-call detection and removal: Detects and removes erroneous base calls identified by elevated motif-associated error rates.
• Adapter sequence removal: Identifies and removes contaminating adapter sequences from reads.
• Barcode swapping mitigation: Detects and mitigates barcode swapping in pooled library sequencing.
• Read-end error identification and tracking: Identifies elevated error rates at read ends and tracks propagation of erroneous base calls.
• Empirically based algorithms: Implements novel, empirically driven algorithms specific to NGS data quality filtering.
• Validation by read compression: Uses read compression rates as an unbiased metric to validate algorithm performance.
• Cross-platform applicability: Concepts and algorithms are applicable across a wide range of NGS protocols and platforms.

Scientific Applications:

• NGS read quality filtering: Improve downstream analyses by removing erroneous bases and adapter contamination from NGS datasets.
• Barcode-swapped library correction: Reduce misassignment effects in pooled library sequencing caused by barcode swapping.
• Objective per-base quality assessment: Provide empirical error estimates to complement platform-generated Phred values that may overestimate quality.
• Method validation: Evaluate and validate filtering effectiveness using read compression rates as an unbiased metric.
• Broad sequencing protocol support: Apply empirical quality-filtering approaches across diverse NGS protocols and sequencing platforms.

Methodology:

Applies empirically based algorithms that leverage known sequence motifs to estimate per-base error rates, detect and remove erroneous base calls and contaminating adapter sequences, identify elevated error rates at read ends and track propagation of erroneous bases; algorithm performance was validated using read compression rates.