HapSolo

HapSolo refines diploid genome assemblies by identifying and removing secondary haplotigs (alternative contigs representing highly heterozygous regions) to produce a more accurate primary assembly.

Key Features:

• Identification of Secondary Contigs: Uses multiple pairwise contig alignment metrics to detect secondary haplotigs and distinguish them from primary contigs.
• Evaluation Using BUSCO Scores: Assesses candidate primary assemblies with Benchmarking Universal Single-Copy Orthologs (BUSCO) scores to evaluate completeness and gene content.
• Cost Function Optimization: Applies a customizable cost function that by default considers the numbers of missing, duplicated, and single-copy BUSCO genes to rank candidate assemblies.
• Hill Climbing Algorithm: Performs hill-climbing optimization over thousands of candidate assemblies to minimize the defined cost function.
• Improved Assembly Metrics: Demonstrated improvements in contig N50 (e.g., ~35% improvement for Chardonnay grape, Vitis vinifera) and reported enhancements for mosquito (Anopheles funestus) and thorny skate (Amblyraja radiata).
• Enhanced Scaffolding with Hi-C Data: Shown to increase downstream scaffolding performance with Hi-C data (e.g., proportion of the Chardonnay genome in the largest 19 scaffolds increased from 52% to ~84%).

Scientific Applications:

• Diploid genome assembly refinement: Produces a refined primary assembly for downstream analyses in variant calling, comparative genomics, and other genomic studies.
• Scaffolding enhancement: Improves scaffolding outcomes when combined with Hi-C data by reducing redundancy from secondary haplotigs.

Methodology:

Computes multiple pairwise contig alignment metrics to identify secondary contigs, evaluates candidate assemblies with BUSCO scores, applies a customizable cost function (defaulting to missing, duplicated, and single BUSCO counts), and uses hill-climbing over thousands of candidate assemblies to minimize that cost.