In supplemental file of ChIPseeker paper, I compare distances to TSS reported by several ChIP annotation software, including ChIPseeker, ChIPpeakAnno, HOMER and PeakAnalyzer.
Although I found that the chromStart positions in HOMER output have a +1 shift compare to other software, I did not realize this issue since all other software are consistent.
ChIPseeker is an R package for ChIP peak annotation, comparison and visualization.
We have implemented several visualization methods, including vennpie that was designed for viewing annotation overlap as shown below:
GOSemSim was first implemented in 2008 and published in Bioinformatics in 2010. It’s now a mature package with no bugs found in the past half year. Only vignette and Information content data were updated.
Almost all annotation software calculate the distance of a peak to the nearest TSS and assign the peak to that gene. This can be misleading, as binding sites might be located between two start sites of different genes or hit different genes which have the same TSS location in the genome.
The function annotatePeak provides option to assign genes with a max distance cutoff and all genes within this distance were reported for each peak.
After two weeks developed, I have added/updated some plot functions in ChIPseeker (version >=1.0.1).
> files=getSampleFiles()
> peak=readPeakFile(files[[4]])
> peak
GRanges object with 1331 ranges and 2 metadata columns:
seqnames ranges strand | V4 V5
|
[1] chr1 [ 815092, 817883] * | MACS_peak_1 295.76
[2] chr1 [1243287, 1244338] * | MACS_peak_2 63.19
[3] chr1 [2979976, 2981228] * | MACS_peak_3 100.16
[4] chr1 [3566181, 3567876] * | MACS_peak_4 558.89
[5] chr1 [3816545, 3818111] * | MACS_peak_5 57.57
... ... ... ... ... ... ...
[1327] chrX [135244782, 135245821] * | MACS_peak_1327 55.54
[1328] chrX [139171963, 139173506] * | MACS_peak_1328 270.19
[1329] chrX [139583953, 139586126] * | MACS_peak_1329 918.73
[1330] chrX [139592001, 139593238] * | MACS_peak_1330 210.88
[1331] chrY [ 13845133, 13845777] * | MACS_peak_1331 58.39
---
seqlengths:
chr1 chr10 chr11 chr12 chr13 chr14 ... chr6 chr7 chr8 chr9 chrX chrY
NA NA NA NA NA NA ... NA NA NA NA NA NA
> covplot(peak, weightCol="V5")
ChIPpeakAnno WAS the only R package for ChIP peak annotation. I used it for annotating peak in my recent study.
I found it does not consider the strand information of genes. I reported the bug to the authors, but they are reluctant to change.
So I decided to develop my own package, ChIPseeker, and it’s now available in Bioconductor.
I used R package ChIPpeakAnno for annotating peaks, and found that it handle the DNA strand in the wrong way. Maybe the developers were from the computer science but not biology background.
> require(ChIPpeakAnno)
> packageVersion("ChIPpeakAnno")
[1] '2.10.0'
> peak <- RangedData(space="chr1", IRanges(24736757, 24737528))
> data(TSS.human.GRCh37)
> ap <- annotatePeakInBatch(peak, Annotation=TSS.human.GRCh37)
> ap
RangedData with 1 row and 9 value columns across 1 space
space ranges | peak strand
|
1 ENSG00000001461 1 [24736757, 24737528] | 1 +
feature start_position end_position insideFeature
1 ENSG00000001461 ENSG00000001461 24742284 24799466 upstream
distancetoFeature shortestDistance fromOverlappingOrNearest
1 ENSG00000001461 -5527 4756 NearestStart
