RFfiller: a robust and fast statistical algorithm for gap filling in draft genomes.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Genome assembly projects frequently encounter gaps or unknown sequences, hindering complete genome reconstruction.
• Existing computational gap-filling methods often suffer from high error rates and mis-assemblies, particularly in repetitive regions.
• Many sequence reads from repeat-related gaps are overlooked by current gap-filling approaches.

Purpose of the Study:

• To present a fast, reliable, and statistically driven algorithm for accurate gap closure in draft genomes.
• To improve the dependability and correctness of de novo genome assemblies.
• To address the limitations of existing methods in handling repeat-related gaps.

Main Methods:

• Developed RFfiller, a statistical algorithm utilizing alignment statistics between scaffolds, contigs, and paired-end reads.
• Employed a Markov chain to assign contigs or long reads to scaffold gap regions for accurate closure.
• Implemented a method to search for valid overlaps in repeat regions and generate transition tables for statistical sequence reconstruction.

Main Results:

• RFfiller demonstrated superior gap-closing accuracy compared to other publicly available tools across diverse genomic datasets.
• Assembly benchmarks confirmed RFfiller's efficiency in filling gaps, with notable effectiveness for longer gaps.
• The algorithm successfully reconstructed missing sequences, outperforming existing gap-closing tools.

Conclusions:

• RFfiller provides an accurate and efficient solution for gap filling in de novo genome assembly.
• The algorithm is particularly effective for closing large gaps and handling complex genomic regions.
• RFfiller represents a significant advancement in improving the quality and completeness of genome assemblies.