HapCHAT

HapCHAT assembles diploid haplotypes from mapped long sequencing reads by leveraging long-range read information and supporting high-coverage datasets (up to 60×) to improve allele assignment accuracy.

Key Features:

• Long-range information utilization: Leverages long sequencing reads that span multiple variants to reduce fragmentation and decrease the number of unphased blocks.
• High-coverage error correction: Processes higher coverage datasets to mitigate the high error rates of long reads by distributing errors across genome positions and improving assembly accuracy.
• Dynamic error estimation: Dynamically adapts the estimated number of errors or error thresholds at each variant site to minimize the total number of error corrections.
• Scalability and efficiency: Handles coverages up to 60× and achieves improvements in accuracy and recall while maintaining lower runtime requirements compared to existing approaches.

Scientific Applications:

• Haplotype-resolved variant analysis: Produces more contiguous haplotypes for assigning alleles across phased genomic regions in genetic variation studies.
• Population genetics: Supports population-scale analyses that require accurate phasing across individuals.
• Complex trait studies: Enables improved interpretation of haplotype structure in analyses of complex trait genetics.
• Personalized medicine: Facilitates more accurate individual-level haplotype information relevant to clinical and translational genetics.

Methodology:

The algorithm assigns alleles from mapped sequencing reads to the two haplotypes, extends previous approaches to accommodate long reads and high-coverage data, and dynamically adjusts per-site error estimates to minimize corrections and reduce haplotype fragmentation.