HiPhase

HiPhase phases variants in diploid organisms using PacBio HiFi long-read sequencing to assign alleles at heterozygous loci to haplotypes for analyses of genetic variation and inheritance.

Key Features:

• Joint Phasing of Multiple Variant Types: Performs joint phasing of single nucleotide variants (SNVs), insertions/deletions (indels), structural variants, and tandem repeat variants.
• Dual Mode Allele Assignment: Employs a dual mode allele assignment strategy to enhance detection and assignment of large and complex variants.
• A* Algorithm Application: Applies an A* search algorithm to optimize haplotype construction and improve phasing efficiency and accuracy.
• Phase Block Spanning Logic: Implements logic to allow phase blocks to span breaks caused by alignment issues around reference gaps and homozygous deletions.
• Performance Metrics: Reported average phase block NG50 of 480 kb with 929 switchflip errors and phasing of 93.8% of genes.
• Concurrent Processing: Supports multi-threading, gathers statistics during execution, and generates concurrent phased alignment outputs.

Scientific Applications:

• Genetic Association Studies: Provides detailed haplotype information to support interpretation of allele-specific effects in association analyses.
• Population Genetics: Enables analysis of phased variant patterns across individuals for population-level haplotype structure and diversity studies.
• Evolutionary Biology: Facilitates examination of haplotype-resolved structural and repeat variation relevant to evolutionary inference.

Methodology:

Leverages PacBio HiFi long reads to call and phase SNVs, indels, structural variants, and tandem repeats; uses a dual mode allele assignment and an A* algorithm for haplotype construction; implements logic to span alignment breaks from reference gaps and homozygous deletions; and supports multi-threading while collecting statistics and emitting phased alignments.