tangles

A script to generate tangle plots of read-mapping positions with circular references

Tangle plots are a way of visualising where paired reads map to a cicular reference. This reference may represent a circular genome or a tandem-repetitive DNA sequence. Tangle plots were used by Becher & Nichols (submitted) to visualise paired reads mapping to the mitochondrial genome in the grasshopper Podisma pedestris. Many read pairs mapped further apart than expected given their insert size, indicating the presence of rearranged nuclear copies of the mitochondrial DNA (Numts).

Requirements

Tangle plots can be generated with R base graphics (we used R version 3.4.4). The criclize package is recommended for annotations.

To generate the input file, you will need bwa, samtools, and the command line utilities cut and gzip:

1. Read mapping should be carried out in single-end mode because, when mapping paired reads, bwa tries to estimate the insert size from the data and will preferentially place reads at a “reasonable-looking” distance. The output should be a BAM file. Example: bwa mem -t <no of threads> <reference> <(zcat <read files>) | samtools view -F 4 -b -o <output.bam>
2. The BAM file with the mapping results needs to be sorted by read name: samtools sort -n -@ <no of threads> -o <output.bam> <input.bam>
3. From the sorted BAM file, read names, positions, and mapping qualities can be extracted with samtools and can be processed with cut utility: samtools view <infile> | cut -f 1,4,14 | gzip > <outfile.gz>
4. The actual plots are generated interactively in R with the script tangles.R. The script contains code to re-plot tangle plots in the Becher & Nichols (submitted, figure 5).

Tangles coloured by average mapping quality:

Tangle plot annotated with circlize: