Chip-seq peak annotation [Chipseeker]

Question

[anu014]

Hi all!

I was dealing with this package called 'Chipseeker' . I tried annotating a chip-seq peak calls fro Mycobacterium tuberculosis. As there is no 'org.Mtu.eg.db' package for annotation & 'TxDB' can be used in Chipseeker, so I created TxDB package & annotated as follows :

library(GenomicFeatures) txdb <- makeTxDbFromGFF("./m.tub1.gff3",circ_seqs=character(), organism="Mycobacterium tuberculosis") peakAnno <- annotatePeak(peak1, tssRegion=c(-3000, 3000), TxDb=txdb, genomicAnnotationPriority = c("Promoter", "5UTR", "3UTR", "Exon", "Intron", "Downstream", "Intergenic"))

Note: GFF was downloaded from ftp://ftp.ensemblgenomes.org/pub/bacteria/release-35/gff3/bacteria_0_collection/mycobacterium_tuberculosis_h37rv/Mycobacterium_tuberculosis_h37rv.ASM19595v2.35.gff3.gz

The chip-seq was done for 'Gyrase' but annotations from Chipseeker gave mostly 'promoters' . Is is correct or am I suppose to add some more steps? Please help me.

Thanking you in advance, Anupriya

[Sinji]

Check your ChIP-seq data on IGV and look at a couple of examples of peak calls. Do they mostly fall in the promoter region? If so, then the annotation is probably correct.

Answer 1

[Guangchuang Yu]

For virus/bacteria, tssRegion=c(-3000, 3000) is too large.

[anu014]

Thank you Guangchuang for the reply :) So how can I optimise the parameter for 'bacteria'? And do parameter 'genomicAnnotationPriority' affects the annotation process in a way that Chipseeker will try to fit promoters first then rest of the types of annotation?

[Guangchuang Yu]

I don't know how to optimise the tssRegion parameter. c(-1000, 1000) or c(-3000, 3000) are commonly used for human. You may need to checkout some references for your species.

Yes, genomicAnnotationPriority parameter will affect the priority of genomic annotation and you can set lower priority of promoter annotation.

[anu014]

Can I use the same range for Mouse too? as it's almost similar to human..

[Guangchuang Yu]

yes. mouse is similar to huaman.

[anu014]

Okay. I'll be using same range for mouse now. Thank you Guangchuang.