Chip-Seq For Mitochondrial Genome

Question

[kanwarjag]

I have a chipseq data where I am looking for mitochondrial genes differential binding. The samples are for human cell lines. My question is if I align reads with hg19 genome which as such will have mitochondrial genome will that give me a correct representation or I have to specifically take mitochondrial genome as reference for using any aligner

Thanks

Answer 1

[Sukhi Singh]

EIther way, you will get the information.

Map all reads and filter the Bed/Bam file for chrM or just subset the reference fasta index before mapping for chrM and then map and proceed with downstream analysis. The former will take more time, depending on the file size.

[epigene]

mitochondrial DNA is circular, does some precautions/procedures need to be taken to deal with that? I assume almost all aligners take linear genome as reference?

[Sukhi Singh]

Oh, I never thought about that, have a no idea. People at BWA/Bowtie support could answer.
There are papers for circular genome mappings though.
http://www.dcs.kcl.ac.uk/staff/csi/publications/ITAB_2010.pdf

[Sukhi Singh]

Hmm, being discussed here, 2 strategies

1. Either add the beginning of the genome/chromosome to the end for the length of the reads, so you don't loose anything.
2. With Bowtie, use -M 1 so that the read is assigned randomly. (quite not clear to me)
3. Use NovoAlign

http://www.researchgate.net/post/NGS_sequencing_issue_When_the_regerence_genome_is_circular