FALCON-Phase

FALCON-Phase is an open-source haplotype phasing tool developed by Phase Genomics that combines Hi-C data with long-read sequencing to produce fully phased diploid genome assemblies. It is designed for researchers working with diploid organisms who need to resolve paternal and maternal haplotypes from a single analysis.

The tool works by examining primary contigs and alternate haplotigs in the context of Hi-C data. Because sister chromatids are independent DNA molecules in the nucleus, they generate independent Hi-C profiles that can be used to identify which heterozygous sequences originated on the same molecule. FALCON-Phase applies a graph partitioning algorithm to detect and correct likely phase switch errors, achieving greater than 96% contig phasing accuracy in known-truth, pedigree-based benchmarks. When used together with Phase Genomics' Proximo scaffolding platform, FALCON-Phase can extend phase blocks to the chromosome scale, delivering two complete, true-phased chromosome sets — both the paternal and maternal genomes — from a single analysis.

Key Features and Capabilities

• Described as the only Hi-C based phasing software available
• Achieves greater than 96% contig phasing accuracy in pedigree-based benchmarks
• Integrates with Proximo to produce complete, fully phased diploid scaffolds at chromosome scale
• Compatible with PacBio's FALCON assembly stack, or with any other long-read technology provided that purge_haplotigs is run
• Accepts input from both FALCON-Unzip and purge_haplotigs workflows

Analysis Workflow — Inputs

• Primary contigs (p_ctgs) and alternate haplotigs (h_ctgs) in FASTA format, produced by either FALCON-Unzip or purge_haplotigs
• Hi-C read pairs in FASTQ format

Analysis Workflow — Outputs

• Phased contigs in two available formats:
  • Pseudohap format — each phase is a complete set of contigs (FASTA)
  • Unzip format — primary contigs and alternate haplotigs with phase switch errors corrected (FASTA)
• Juicebox files (HIC and ASSEMBLY formats)
• Aligned Hi-C reads (BAM format)
• QC reports (PDF and JPG formats)
• Various intermediate files and reports
• When combined with Proximo, one full set of Proximo outputs for each phase

Access and Availability

• FALCON-Phase is open-source and available via the Phase Genomics GitHub page at github.com/phasegenomics
• Phase Genomics can be contacted directly to design or initiate a FALCON-Phase project with support

FALCON-Phase has been applied to produce highly contiguous diploid human genome assemblies by combining PacBio long-read and Hi-C data. It is part of Phase Genomics' broader suite of cloud-based bioinformatic platforms that analyze and integrate proximity ligation data for applications in cytogenomics, metagenomics, and epigenomics.