Phasing nanopore genome assembly by integrating heterozygous variations and Hi-C data

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Haplotype-resolved genome assemblies are crucial for genomics, medicine, and pangenomics.
• Hi-C data offers an advantageous approach for generating these assemblies due to its availability.
• Current methods struggle to balance accuracy and the number of informative Hi-C alignments by relying solely on mapping quality.

Purpose of the Study:

• To introduce Diphase, a novel phasing tool designed to improve the accuracy of haplotype-resolved genome assemblies.
• To leverage heterozygous variations alongside mapping quality for more precise filtering of Hi-C alignments.
• To enhance the extension of primary and alternate assemblies.

Main Methods:

• Developed Diphase, a computational tool that utilizes heterozygous variations to identify informative Hi-C alignments.
• Integrated mapping quality and heterozygous variation data for filtering uninformative alignments.
• Performed comparative analysis of Diphase against FALCON-Phase and GFAse using human datasets.

Main Results:

• Diphase demonstrated superior performance in comparative analyses on human datasets.
• Achieved a longer phased block N50, indicating improved assembly contiguity.
• Exhibited higher phasing accuracy and a lower Hamming error rate compared to existing tools.

Conclusions:

• Diphase effectively enhances the accuracy and contiguity of haplotype-resolved genome assemblies.
• The integration of heterozygous variations offers a robust strategy for filtering Hi-C data.
• Diphase represents a significant advancement in phasing tools for genomic research.