Hi,

I am trying to annotate some peaks from a RIP seq experiment. They are supposed to overlap annotated genes.

My problem is that the assigned gene in ChIPSeeker for some of my peaks does not match the gene which is annotated in the annotation column of the returned csAnno object and which is actually overlapping the peak.

Here my example:

library(ChIPseeker)
library(GenomicFeatures)
library(biomaRt)

txdb <- makeTxDbFromBiomart(biomart="ENSEMBL_MART_ENSEMBL",
                              dataset="dmelanogaster_gene_ensembl",
                              host="mar2015.archive.ensembl.org")

test_gene2 <- GRanges(c("3L"), IRanges(c(8597160), end=c(8597371)), strand=c("+"))

annotated <- annotatePeak(test_gene2, TxDb=txdb, level="transcript", tssRegion=c(-200,200))

> as.data.frame(annotated)

as.data.frame(annotated)
  seqnames   start     end width strand                                  annotation geneChr geneStart geneEnd geneLength geneStrand      geneId transcriptId distanceToTSS
1       3L 8597160 8597371   212      + Exon (FBtr0340707/FBgn0262508, exon 8 of 8)      3L   8601006 8602392       1387          + FBgn0017579  FBtr0076633         -3846

The gene which is below annotation is Exon (FBtr0340707/FBgn0262508, exon 8 of 8)  but the gene reported is FBgn0017579 which is the gene with the nearest TSS but actually not the overlapping one. I did not find an option to change that and for me it would make sense to actually report the gene the peak is overlapping with instead of the nearest TSS gene.  Especially if we are looking at marks which may be located on or near 3'UTRs of genes.

Is there a way to change this behaviour? Is this a feature or a bug?

Thank you very much for any comments.

Regards

Nastasja Kreim

pls refer to the issue: https://github.com/GuangchuangYu/ChIPseeker/issues/12

Hi Yu,

thanks for the fast answer.

It would be great to mention in the description of the function (annotatePeak) that the gene reported is not the nearest gene but the gene with the nearest TSS. In my example the peak is contained within a gene which should be the nearest but is not reported as such since its TSS is farther away from the location in the 3'UTR than the TSS of the next gene.

Or it would be great to have a switch changing this behaviour.

Regards

Nastasja