DENTIST-using long reads for closing assembly gaps at high accuracy.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Short-read sequencing technologies often result in fragmented genome assemblies with gaps.
• Existing gap-closing methods using long reads can introduce inaccuracies.
• Improving genome assembly contiguity and completeness is crucial for genomic research.

Purpose of the Study:

• To develop a highly accurate and automated pipeline for closing gaps in genome assemblies using long reads.
• To enhance the contiguity and completeness of fragmented genome assemblies.
• To provide a reliable method for accurate gap-filling in genomic data.

Main Methods:

• Developed DENTIST, an automated pipeline for gap closure in short-read assemblies with long, error-prone reads.
• Implemented comprehensive determination of repetitive regions for unambiguous long-read alignment.
• Integrated a consensus sequence computation for high base accuracy and validated closed gaps.
• Utilized realistic benchmark datasets for Drosophila, Arabidopsis, hummingbird, and human genomes.

Main Results:

• DENTIST demonstrated substantially higher accuracy in closing assembly gaps compared to previous methods.
• The method achieved high sensitivity in gap closure across various genome sizes.
• Validated accuracy of closed gaps using simulated and real PacBio continuous long reads.
• DENTIST consistently improved genome assembly contiguity and completeness.

Conclusions:

• DENTIST offers an accurate and automated solution for improving genome assembly contiguity and completeness using long reads.
• The pipeline provides a significant advancement over existing gap-closing techniques.
• Source code and benchmark datasets are publicly available for reproducibility and future research.