GATK ReadBackedPhasing short haplotypes

[gmdouglas]

I am trying to sort polymorphisms to haplotypes on both homeologs of a plant species. This is the command I am using for chromosome 1 of one individual:

java -jar GenomeAnalysisTK-1.6-5-g557da77/GenomeAnalysisTK.jar -T ReadBackedPhasing -R reference.fasta -L scaffold_1:1-5000000 -I file.bam --variant file.vcf -o phased_file.vcf 2>error_file.err &

Firstly, the program runs extremely slowly, does anyone know a way to speed up the process?

Secondly, the majority of haplotypes generated are very short. I need larger haplotypes so I can assemble the homeologs using haplotypes from a number of individuals. Has anyone encountered a similar problem?

Thanks in advance!

[SeekAnswers]

Just out of curiosity, did you write a parser script to generate the haplotypes from the GATK VCF output?

[gmdouglas]

Yep, it turns out I was counting the number of variants per haplotype instead of the entire haplotype length.