TGS-GapCloser

TGS-GapCloser closes gaps in draft genome assemblies using low-depth (≥10X) single-molecule third-generation sequencing (TGS) long reads without prior error correction to improve assembly contiguity and single-base accuracy.

Key Features:

• Utilization of Low-Depth Long Reads: Employs low-depth (≥10X) single-molecule TGS long reads without prior error correction to leverage medium/long-range information for gap filling.
• Gap identification and correction: Distinguishes gap regions from alignments of long reads against original scaffolds, selectively corrects candidate regions, and assigns optimal sequences to each identified gap.
• Assembly metric improvement: Improves contig N50 of draft assemblies by ~25-fold on average and updates over 90% of gaps with a positive predictive value of 93.96%.
• Single-base accuracy: Produces assemblies with Q50 (99.999%) single-base accuracy with an 11.8% decrement relative to inputs.
• Computational efficiency: Executes approximately 29-fold faster than mainstream gap-closing tools.
• Gene completeness: Increases BUSCO complete gene counts by 3.4%–13.1% compared to other methods.
• Ultra-large genome application: Demonstrated closure of 71.6% of gaps in the ginkgo genome (~12 Gb).
• Robustness to multi-alignments and high error rates: Addresses challenges from high error rates and multi-alignments inherent to raw long reads.
• Validation capability: Supports validation of inserted or merged gap sequences using NGS reads and reference genomes.

Scientific Applications:

• Genome assembly improvement: Enhances contiguity and single-base accuracy of draft genome assemblies for downstream analyses.
• Gene annotation and variant analysis: Facilitates more reliable gene annotation and structural variant calling by reducing assembly gaps.
• Comparative and evolutionary genomics: Supports ontogeny, phylogeny, and evolutionary studies by providing more complete assemblies.
• Sequence validation: Enables validation of filled or merged gap sequences using NGS reads and reference genomes.

Methodology:

Aligns low-depth (≥10X) single-molecule TGS long reads to original scaffolds, distinguishes gap regions from read alignments, selectively error-corrects candidate regions, assigns optimal sequences to each gap, and supports validation with NGS reads and reference genomes.