Recently I recevied the result of WGS from illumina, and I see each bam file in one sample has up to 100G!! I want to ask why the size is so big. With so big bam files, if i need to call mutations between samples, can I run samtools mpileup directly on that?
3 answers
The size is big because modern sequencers are very good at producing tons and tons of sequence data!
Yes, you can run samtools mpileup on it. As the samtools website recommends:
samtools mpileup -uf ref.fa aln.bam | bcftools view -bvcg - > var.raw.bcfbcftools view var.raw.bcf | vcfutils.pl varFilter -D100 > var.flt.vcf
As a further example of expected data produced and corresponding file sizes... In our group, three lanes of an Illumina HiSeq sequencer using v3 chemistry will produce enough data for 30-35X haploid coverage of a human genome (approximately 100-120 Gbp of sequence data). When this data is aligned with BWA, a BAM file of 85-95G will be typical with a BAM index file of 7-10M.
Most probably, the each sample's material was divided and run on more than 3-4 lanes of a high throughout sequencer like Hiseq 2000, acting like technical replicates and then merged back later on. In case of ChIP-Seq if one singleplex's with low material, there are lot of duplicates observed.
Log in to answer this question.