Bisulfite_tools

Pipeline for alignment of Illumina bisulfite sequencing data and associated tools

Author: Aaron Statham (a.statham@garvan.org.au)

Usage:

prepare_genome.sh {Genome}

Execute in the directory containing the genome for aligning to as a single multifasta file with the name {Genome}.fa

To build an index for the human genome:

# Download and extract the genome
wget -O - http://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/chromFa.tar.gz | tar -zxv

# Concatenate into a multifasta file
cat *.fa > hg19.fa
rm chr*.fa

# Build index
prepare_genome.sh hg19

align_lane_(PE/SE).sh {Project} {Genome}

Master script for aligning a lane of data. Raw data must be named in the format:

input/{Project}_R1.fastq.gz input/{Project}_R2.fastq.gz

or for a single end experiment:

input/{Project}.fastq.gz

Aligned data, methylation calls and run metrics will be put in a folder named "output".

The pipeline qsubs the following steps

• prep_reads.sh - Adaptor and quality score trimming, fastqc and splitting reads into 20 chunks for alignment in parallel
• align_(PE/SE).sh - Actual alignment is performed by bismark, SAM header is reordered and reads are sorted. Run in parallel as an array job.
• process_lane.sh - The 20 chunks are merged, duplicates are removed, some metrics are collected by Bisulfite_stats.sh, bismark_methylation_extractor is run and then CpG calls are imported into an R object.
• summarize_lane.sh - Alignment metrics are collated.