Explore ChIP-seq alignments from a bam file, MACS2 narrowPeaks, conservation, H3K4me3 methylation and motif matching
2019-11-01
Package
igvR 1.6.1
Contents
- 1 Overview
- 2 Load the libraries we need
- 3 Create a GRanges object specifying our region of interest, obtaining a list (chrom, start, end, string) from igv
- 4 Load narrow peaks previously called by MACS2 from the bam file
- 5 Match the CTCF motif against sequence in the currently displayed region
- 6 Use matchPWM from the Biostrings package to match the motif on both strands
- 7 Set up motif logo viewing
- 8 Render the motif matching in two ways, with two tracks: one for annotation, one for score
- 9 Is there any reason to believe these motif matches are functional?
- 10 Add Methylation Tracks
- 11 Some tentative conclusions
- 12 Session Info
1 Overview
The igvR package provides easy programmatic access in R to the web-based javascript library igv.js. R’s access to data, and operations upon data, are complemented by igv’s richly interactive web browser interface to display and explore genomic and epigentic data.
In this vignette we present a somewhat contrived ChIP-seq study - contrived in that it is not organized around an actual research problem. Instead, this vignette demonstrates methods you would likely use to do visual QC and exploratory data analysis a ChIP-seq experiment.
We begin with ENCODE data of CTCF binding on chromosome 3, on the hg19 reference genome, in the vicinity of the GATA2 gene. We will
- display a 252 kb region of that genome
- load and display a GenomicAlignments object from an indexed bam file covering that region
- load and display a narrow peaks bedGraph track, precalculated by MACS2
- Use the Biostrings::matchPWM method to identify regions in the displayed sequence which, with at least 80% fidelity, match the CTCF motif, Jaspar 2018’s MA0138.1, from MotifDb
- Display that motif logo in a browser popup.
- load sequence conservation data from the Bioconductor annotation package phastCons100way.UCSC.hg19, looking for the intersection of ChIP peaks, motif match and conserved sequence, all of which together suggest that at such locations functional TF binding occurs.
The two images below illustrate the final state of igv in your browser at the conclusion of the code found below. The first view spans 35kb:
The second spans less than 1200 bases.