Optimal haplotype assembly of polyploids

Summary

Haplotype assembly of polyploids is an open issue in plant genomics. Recent experimental studies have shown that available methods are not delivering satisfying results in practice. We investigated integer linear programming (ILP) models to assemble optimal haplotypes in organisms with ploidy up to 8. Our tool is available for non-commercial purposes subject to a license agreement.

Included are also example input files and programs to evaluate phasing results for MEC or haplotyping distance as described in the reference publication.

Reference

Siragusa, E., Haiminen, N., Finkers, R., Visser, R. and Parida, L. Haplotype assembly of autotetraploid potato using integer linear programming. Bioinformatics, 2019. https://academic.oup.com/bioinformatics/advance-article/doi/10.1093/bioinformatics/btz060/5301312

Contact

For questions or comments, feel free to contact: Laxmi Parida <parida@us.ibm.com>

Download

The latest binaries for Linux can be downloaded from the GitHub releases page.

Usage

Our tool takes in input one BAM/SAM file and one single-sample VCF file. It produces in output one single-sample VCF file that encodes phased variants. All files must be sorted by genomic coordinates.

For instance, to assemble tetraploid haplotypes in a genomic region, execute the following command line:

$ haplotype_assembler --ploidy 4 --input-bam INPUT.bam --input-vcf INPUT.vcf --output-vcf OUTPUT.vcf

Option --dosages determines how strictly allelic dosages are regarded in the assembly:

• exact corresponds to the fully constrained model in [1].
• hard corresponds to model cMFR in [1].
• soft corresponds to model aMFR in [1].

Detailed usage below by running $ haplotype-assembler --help

---

SYNOPSIS haplotype-assembler [OPTIONS] -ib <BAM FILE IN> -iv <VCF FILE IN>

DESCRIPTION

OPTIONS

-h, --help	Display the help message.
--version-check BOOL
	Turn this option off to disable version update notifications of the application. One of 1, ON, TRUE, T, YES, 0, OFF, FALSE, F, and NO. Default: 1.
--version	Display version information.
-v, --verbose	Output verbose information to standard error.
-ib, --input-bam INPUT_FILE
	Input SAM/BAM file. Valid filetypes are: .sam[.*] and .bam, where * is any of the following extensions: gz and bgzf for transparent (de)compression.
-iv, --input-vcf INPUT_FILE
	Input VCF file. Valid filetype is: .vcf[.*], where * is any of the following extensions: gz and bgzf for transparent (de)compression.
-ov, --output-vcf OUTPUT_FILE
	Output VCF file. Default: output to standard output. Valid filetype is: .vcf[.*], where * is any of the following extensions: gz and bgzf for transparent (de)compression.
-oh, --output-het
	Output only heterozygous variants in the VCF file.
-p, --ploidy INTEGER
	Organism ploidy. In range [1..8]. Default: 4.
-d, --dosages STRING
	Use dosages from the input VCF file as hard or soft constraints, or ignore them. One of ignore, exact, hard, and soft. Default: exact.
-c, --coverage INTEGER
	Minimum coverage per allele. In range [1..20]. Default: 1.
-t, --timeout INTEGER
	Timeout in seconds. In range [0..3600]. Default: 0.