Skip to content

Latest commit

 

History

History
42 lines (26 loc) · 1.7 KB

pipeline.md

File metadata and controls

42 lines (26 loc) · 1.7 KB
Raw

Requirements:

  1. HapCUT2: https://github.com/vibansal/HapCUT2
  2. SHAPEIT2: https://mathgen.stats.ox.ac.uk/genetics_software/shapeit/shapeit.html
  3. 1000 Genomes reference panel haplotypes: https://mathgen.stats.ox.ac.uk/impute/1000GP_Phase3.html
  4. BAM file and VCF file for individual to be phased

Notes:

The code was developed and tested using python2.7. Currently, it doesn't work for python versions 3 or more.

'VCF' refers to the path to the input VCF file

Steps:

  1. Run extractHAIRS on the BAM and VCF file to obtain Hi-C fragments (see HapCUT2 instructions for how to do this, use option --hic 1)

  2. Run SHAPEIT2 on the VCF file using the 1000 Genomes reference panel:

shapeit -check --input-vcf VCF -R reference_panel.hap.gz reference_panel.legend.gz reference_panel.samples --output-log VCF.log
shapeit --input-vcf VCF -R reference_panel.hap.gz reference_panel.legend.gz reference_panel.samples -M genetic_map.txt --output-graph VCF.graph --exclude-snp VCF.check.snp.strand.exclude

The files reference_panel.hap.gz, reference_panel.legend.gz and reference_panel.samples (for each chromosome) can be downloaded from the link above.

  1. use samplehaps.py to sample 'N' haplotype pairs for the individual (we have used N=1000). The program will output the samples to the file VCF.hapsamples where VCF is the path to the input VCF file
python samplehaps.py VCF 1000
  1. use encodereads.py to represent the haplotype samples as "pseudo-reads"
python encodereads.py VCF.hapsamples > sample.pseudo_reads

  1. concatenate the "Hi-C fragments" file and the sample.pseudo_reads to obtain the new "fragment file"

  2. run HapCUT2 on the new "fragment file" to obtain phased haplotypes