Variant calling and analysis

The main steps comprising variant calling and analysis are

• mapping short reads
• calling raw variants
• filtering (really more like 'tagging') variants
• annotating variants

Map short reads

After quality control and/or filtering of reads...

Single reads with bowtie2:

bowtie2 -x /nfs/genomes/sgd_2010/bowtie/sacCer3 A_reads.fq -S A_reads.bt2.sam

Single reads with bwa:

bwa aln /nfs/genomes/sgd_2010/bwa/sacCer3.fa A_reads.fq > A_reads.sai
bwa samse /nfs/genomes/sgd_2010/bwa/sacCer3.fa A_reads.sai  A_reads.fq > A_reads.bwa.sam

Paired-end reads with bowtie2:

bowtie2 -x /nfs/genomes/sgd_2010/bowtie/sacCer3 -1 A_reads.1.fq -2 A_reads.2.fq -S A_reads.1+2.bt2.sam

Paired-end reads with bwa:

bwa mem /nfs/genomes/sgd_2010/bwa/sacCer3.fa A_reads.1.fq A_reads.2.fq > A_reads.1+2.bwa.sam

Convert to BAM, sort, and index [with a custom BaRC script that uses samtools]:

/nfs/BaRC_Public/BaRC_code/Perl/SAM_to_BAM_sort_index/SAM_to_BAM_sort_index.pl A_reads.bwa.sam

Get uniquely mapping reads

samtools view -h A_reads.bt2.bam | grep -v XS:i: | samtools view -bS - > A_reads.bt2.sorted_unique.bam

Call raw variants